1 s2.0 S2590262823000412 Main

Current Research in Biotechnology 6 (2023) 100159
Contents lists available at ScienceDirect
Current Research in Biotechnology

journal homepage: www.elsevier.com/locate/crbiot
An amalgamation of bioinformatics and artificial intelligence for COVID-19

management: From discovery to clinic
Jiao Wang a, b, 1, Vivek Chavda c, *, 1, Riddhi Prajapati c, Anjali Bedse d, Jinita Patel c, Sagar Popat c,
Gargi Jogi c, Lakshmi Vineela Nalla e, Keshava Jetha c, f, Bairong Shen a, *, Rajeev K. Singla a, g, *
a
Joint Laboratory of Artificial Intelligence for Critical Care Medicine, Department of Critical Care Medicine and Institutes for Systems Genetics, Frontiers Science Center
for Disease-related Molecular Network, West China Hospital, Sichuan University, Chengdu, China.
b
Department of Computer Science and Information Technology, University of A Coruña, A Coruña, Spain
c
Department of Pharmaceutics and Pharmaceutical Technology, L.M. College of Pharmacy, Ahmedabad 380009, Gujarat, India
d
Department of Pharmaceutics, K K Wagh College of Pharmacy, Nasik, Affiliated to Dr. Babasaheb Ambedkar Technological University, Lonere, Raigad, Maharashtra,
India
e
Department of Pharmacology, GITAM School of Pharmacy, GITAM (Deemed to be University), Rushikonda, Visakhapatnam, Andhra Pradesh 530045, India
f
Pharmacy Section, Gujarat Technological University, Chandkheda, Ahmedabad, Gujarat 380005, India
g
School of Pharmaceutical Sciences, Lovely Professional University, Phagwara, Punjab 144411, India
A R T I C L E I N F O A B S T R A C T
Keywords: The pathogen SARS-CoV-2 has emerged and taken the shape of a global pandemic by causing COVID-19. SARS-
SARS-CoV-2 CoV-2 has a very novel and unique set of genetic makeup that has created a puzzle in biological research.
COVID-19 Therefore, the scientific community has not yet discovered a very effective new treatment or preventive solution.
Bioinformatics
By using various bioinformatics techniques and tools, the decoding of the genomic structure of the virus has been
Next generation sequencing
Genome-wide association study
possible. In COVID-19 research methodologies, next-gene sequencing and computer-aided drug design have been
Drug design implemented to decode this new structure of the SARS-CoV-2. Implementing in silico studies for COVID-19 has
analyzed various evolutionary relationships, sequencing errors, and for determining potential candidates against
the SARS-CoV-2 genes, and that too in a short span. The information derived using various bioinformatics
techniques would fast forward the research speed on the SARS-CoV-2 and provide essential information for
vaccine development, which is essential to ensure the overall betterment of public health. The application of
artificial intelligence (AI) and Machine learning (ML) has provided an attractive niche for the bio-therapeutics
development for COVID-19. This review article describes the application of AI and ML for the therapeutic
management of COVID-19.
Introduction precise COVID data, aiding healthcare professionals in gaining insights

into the actual nature of the virus (Sharma, 2021). The introduction of
In the early phase, Severe acute respiratory syndrome coronavirus 2 genomics, bioinformatics, and artificial intelligence (AI) has made a
(SARS-CoV-2) posed a perplexing challenge as doctors and scientists notable contribution to understanding the pathogenesis of the disease,
struggled to identify effective treatments for the rapidly mutating virus its treatment, and the mechanism by which antimicrobial resistance is
(Chavda et al., 2021). A limited understanding of the virus and its increased to the host’s immune response (Hufsky et al., 2021). To
structure and rapid spread led to uncertainty. Time constraints hindered develop efficacious and innovative prevention strategies and design
the compilation and analysis of medical and scientific observations. The novel treatment approaches, a comprehensive investigation is impera
Swiss Biotech Association developed an AI-based platform called RIS tive to grasp the pathogenic and virological characteristics of SARS-CoV-
KLICK AI to address this issue. This platform successfully gathered 2 (Artificial intelligence–enabled rapid diagnosis of patients with
* Corresponding authors at: Joint Laboratory of Artificial Intelligence for Critical Care Medicine, Department of Critical Care Medicine and Institutes for Systems
Genetics, Frontiers Science Center for Disease-related Molecular Network, West China Hospital, Sichuan University, Chengdu, China (R.K. Singla and B.Shen).
E-mail addresses: [email protected] (V. Chavda), [email protected] (A. Bedse), [email protected] (B. Shen), [email protected],
[email protected] (R.K. Singla).
1
Both authors contributed equally and will be treated as First authors.
https://doi.org/10.1016/j.crbiot.2023.100159
Received 13 July 2023; Received in revised form 29 October 2023; Accepted 10 November 2023
Available online 16 November 2023
2590-2628/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
J. Wang et al. Current Research in Biotechnology 6 (2023) 100159
COVID-19, 2023). AI and biopharmaceutics have found practical ap Bali, 2022). Wang et al., in their systemic review of the application of AI
plications in numerous therapeutic domains, particularly in aiding dis in the context of COVID-19, have discussed different protocols that are
ease diagnosis and prevention. Their integration has facilitated essential for AI tools. The findings have demonstrated the remarkable
advancements in various areas of healthcare, enabling more accurate potential of artificial intelligence across various critical areas. AI has
and timely disease identification and the development of preventive proven its efficacy in achieving high-performance levels in tasks related
measures (Artificial intelligence–enabled rapid diagnosis of patients to diagnosing COVID-19 cases, assessing the prognosis of patients, pre
with COVID-19, 2023). During the difficult time of the pandemic, when dicting the spread of the epidemic, and even aiding in discovering po
the health care resources were limited, the utilization of various AI- tential drugs for the treatment of the virus (Wang et al., 2021).
derived technologies played a crucial role in diagnosis, treatment Esteemed institutions like Johns Hopkins University and the OECD
exploration, and enhancement of public health decision-making, (oecd.ai) have developed collaborative dashboards that provide acces
thereby bolstering the potential to combat the pandemic. Various AI- sible tracking of the virus’s progression. These dashboards integrate
based techniques have been developed, integrating the diagnostic sys real-time news updates with historical data on confirmed coronavirus
tems of COVID-19. This includes techniques such as Recurrent Neural cases, recoveries, and fatalities, offering comprehensive situational
Network (RNN), Long Short-Term Memory (LSTM), Graphic Neural awareness (Arslan and Benke, 2021). The focus should be on population
Network (GNN), and Extreme Learning Machines (ELM) (Chen and See, data collection to ensure accuracy and capability (Fig. 1) (OECD, 2023).
2020). Extensive research endeavors have been dedicated to formu This review paper explores the interdisciplinary intersection of these
lating and implementing effective strategies to tackle the challenges technologies, clarifying their combined capacity to expedite the process
presented by the COVID-19 pandemic (Robson, 2020). of drug discovery, improve treatments, and enhance predictive
When formulating therapeutic strategies, it is crucial to consider the modeling. Our objective is to completely comprehend the trans
virus’s pathogenesis (Cardiovascular Implications of Fatal Outcomes of formational nexus in modern healthcare by highlighting significant
Patients With Coronavirus Disease 2019, 2019). AI and machine breakthroughs, critical studies, and future prospects.
learning (ML) based algorithms have been deployed to integrate infor
mation from chest computed tomography scans, clinical symptoms, Sequencing for COVID-19
exposure history, and laboratory tests, enabling the swift and accurate
diagnosis of individuals with COVID-19 (Artificial Intelligence and Because of the fast evolution of SARS-CoV-2, the sequencing
COVID-19, 2023; Chavda et al., 2023). AI can also help to identify the approach is critical in many ways (Chavda et al., 2022). This section
specific segments or regions within the viral protein sequence. This can summarizes current knowledge of various sequencing approaches to
be crucial as such studies help develop peptide-based vaccines and identify SARS-CoV-2 based on NGS data.
peptidomimetic therapeutics while exploring various drug design pos
sibilities (Managing COVID-19 in resource-limited settings, 2023). Viral whole genome sequencing approach
AI works by amassing vast amounts of data with quick, unvarying
processes and intelligent algorithms. This permits the software to learn The groundwork scheme hatched by the nCoV-2019 ARTIC network
automatically from patterns or options within the knowledge and utilize utilizes preliminary V1 sets to generate an amplicon tiling method
the available resources in the best possible ways for the population. AI throughout the viral genome (Fig. 2) (Wang et al., 2020). cDNA MinION
uses major subfields like ML, Neural Network (NN), and Deep Learning library planning was conducted using the fast barcoding unit following
(DL) (Peiffer-Smadja et al., 2020; Chavda et al., 2023). Such AI tools and manufacturer requirements and then stacked into a MinION stream cell
methods automate the identification of the root cause and often aid (Mishra, 2020). Using Guppy (rendition 3.4.5), the base calling of FAST5
policymakers in drafting proper protocols and guidelines (Chavda et al., documents was carried out, and RAMPART programming (v1.0.5) from
2022). AI text and data processing tools can quickly reveal the virus the ARTIC Network was used to screen sequencing gradually. NanoPlot
history, spread, medicine, management processes, and lessons from 13 and Porechop calculations were used for succession efficiency
earlier epidemics like SARS (2003) and MERS (2012) outbreaks (Chavda scoring, demultiplexing, and connector evacuation. The ARTIC Network
et al., 2022; Huang et al., 2022). It can also be used to identify zones of bioinformatics program was utilized for variant calling, and the peruses
vigorous transmissions and help to restrict rapid transmission. AI offers were developed against the reference strain Wuhan-Hu-1 to build
limitless possibilities for rapid action against spreading and restricting agreement genomes. The genome was sent to the CoV-GLUE asset to
the virus in a limited population. assess the epidemiological connection of SARS-CoV-2 mutations (Hafeez
A lack of surveillance procedures and a weaker healthcare system et al., 2020).
have made the COVID-19 pandemic especially serious. Many more
severely impacted areas have naturally shown an interest in AI tech Metagenomics approach and bioinformatics analysis
nology, expecting them to provide effective and affordable solutions to
medical and societal problems (Piccialli et al., 2021). AI-assisted early The cDNA MinION library preparation followed the manufacturer’s
warning systems can enable the monitoring of patterns across various instructions using the rapid barcoding kit (Wei et al., xxxx). The
medical specialties by analyzing diverse sources such as news articles, resulting library was loaded onto an Oxford MinION flow cell and sub
online content, and multilingual data channels. By generating early jected to sequencing using the MinKNOW software for twenty-four
warnings, these systems can complement and improve the existing hours. Base calling and quality control analysis were performed using
syndromic surveillance and healthcare networks, such as the WHO Early Guppy (version 3.4.5) and NanoPlot (version 1.29.0), respectively.
Warning System and Bluedot (Arslan and Benke, 2021). Additionally, AI Porechop (version 0.2.4) extracted 13 adapters and standard barcodes
techniques can facilitate the identification of viral transmission net from the MinION chip. For precise grouping, the Kaiju platform (version
works and provide insights into the broader economic impacts of out 20) was utilized (White et al., 2017).
breaks. The application of AI to speed up research on COVID-19 vaccine Reads were aligned against the reference strain Wuhan-Hu-1 (Gen
and medication design has been examined. It should be emphasized that Bank accession number MN908947) using the minimap to obtain a
antiviral research is a highly multifaceted and perplexing area of med consensus sequence for the SARS-CoV-2 alignment. Finally, Samtools
icine, and prediction accuracy is crucially dependent on the currently (version 1.9) and Tablet alignment viewer (version 1.19.09.3) were
available data. Because protein research is a discipline that is continu utilized to visualize the aligned sequence (Woo et al., 2005).
ally changing and presenting more protein targets with conceivably
newer processes, even the mechanisms of action of currently available
medications against different kinds of viruses are not clear (Bali and
2
Fig. 1. Comparison between AI and non-AI approaches for COVID-19 Management (created with Biorender.com).
Fig. 2. Sequencing approaches for COVID-19 (created with Biorender.com).
Shotgun metagenomics syndrome, it has been observed that some patients also experience
gastrointestinal (GI) symptoms. Gut microbiota imbalances have previ
Although COVID-19 is primarily associated with respiratory distress ously been associated with respiratory viral infections. In COVID-19
3
patients, there is a reduced presence of butyrate-producing Roseburia have raised global concerns due to potential increased transmissibility,
inulinivorans, while the presence of Paraprevotella species is relatively altered disease severity, or reduced vaccine efficacy (Chavda et al.,
high. These findings suggest that altered gut microbes in COVID-19 2021). Continuous surveillance, genomic sequencing, and adaptable
patients are associated with the severity of the condition (Li et al., techniques are pivotal in the global health response to these dynamic
2021). This strategy can help quicken flare-up examinations and back issues.
improvement of new lab tests (Hafeez et al., 2020).
• Variant of Concern (VOC)
Target enrichment sequencing
Omicron (BA.2): South Africa first detected it in November 2021.
Target Enrichment catches genomic districts of interest by employ G142D, N211I, Δ212, V213G, G339D, S371F, S373P, S375F, T376A,
ing hybridization to target-explicit tests (Groneberg et al., 2005). Whole D405N, R408S, K417N, N440K, S477N, T478K, E484A, Q493R, Q498R,
genome sequencing is considered the gold standard method for N501Y, Y505H, D614G, H655Y, N679K, P681H, N764K, D796Y,
achieving the highest accuracy and reliability in genetic analysis. Q954H, N969K are the spike mutations of interest. It has a high impact
However, it is a more expensive approach. It generates a vast amount of on immunity but a reduced impact on severity (Aleem et al., 2023).
data, which can decrease the frequency of research studies. Target Omicron (BA.4): South Africa first detected it in January 2022.
enrichment methods, particularly hybridization techniques, have revo L452R, F486V, and R493Q are spike mutations of interest. It has a high
lutionized genetic analysis. These methods include molecular inversion impact on immunity but no evidence of severity.
probes, primer extension, and PCR-based approaches. In the context of BA.5: South Africa first detected it in February 2022. L452R, F486V,
combating the COVID-19 pandemic, the preparation of a single-cell RNA and R493Q are spike mutations of interest. It has a high impact on im
library is one notable application of target enrichment. For instance, the munity, but is unclear about its impact on severity (Choi and Smith,
Single Index Kit N Set A was employed to generate CiteSeq libraries. The 2021).
Chromium Single Cell V(D)J Enrichment Kit for Human T and B cells
was utilized to enrich T-cell receptor (TCR) and B-cell receptor (BCR) • Variants of Interest (VOI)
targets. Subsequently, GEX (gene expression) and TCR/BCR libraries
were constructed, involving fragmentation, adapter ligation, and a final Omicron (BA.2.75): India detected it in May 2022. W152R, F157L,
Index PCR. The resulting libraries were quantified using a Qubit HS DNA I210V, G257S, D339H, G446S, N460K, and Q493 (reversion) are spike
assay kit from Life Technologies for database quantification (Huang mutations of interest. It has a high impact on immunity but has no ev
et al., 2004). idence of impact on severity.
Omicron (BQ.1): K444T and N460K are spike mutations of interest. It
Amplicon sequencing has a high impact on immunity but has no evidence of impact on
severity.
Amplicon sequencing, which involves the targeted amplification of XBB Recombinant lineage of BJ.1 (BA.2.10.1.1) and BM.1.1.1
specific RNA or DNA regions, is crucial in studying viral strains with (BA.2.75.3.1.1.1): N460K, F490S are spike mutation of interest (Konings
precision (Wensel et al., 2022). This sequencing method allows for et al., 2021).
identifying genomic variations in a more accurate and detailed manner
compared to standard Nucleic Acid Amplification tests used in COVID- • Variants of Monitoring (VOM)
19 diagnosis. Testing based on targeted sequencing of SARS-CoV-2
demonstrated excellent specificity. It has been observed that using Omicron (BA.2.3.20): K444R, L452M, N460Kare spike mutation of
shorter amplicons (less than 100 base pairs) can maximize sensitivity, interest
surpassing larger amplicons (over 200 base pairs) (McNamara et al., Omicron (BF.7): R346T and F486V are spike mutations of interest.
2020). XBC Delta (21I) / BA.2 recombinant, Omicron spike with Delta back
bone: N440K, and F486P are spike mutations of interest (Esman et al.,
SARS-CoV-2 susceptible organs with high expression of ACE2 determined 2022).
by Next-Generation sequencing
• De-Escalated variants:
RNA sequencing technology has enabled dynamic genome
improvement and qualitative and quantitative alterations in DNA and
RNA. Such alterations have been used for many transcriptomic and Alpha variant (B.1.1.7)
genomic assessments (Ergin et al., 2022). However, RNA sequencing had It was first detected in September 2020 in the United Kingdom. The
many challenges, including limited sequencing depth for low- mutation Number is 7. Defined in S proteins are D1118H, N501Y,
abundance transcripts and ineffaceable influencing of interferential D614G A570D T716I, S982A, and P681H. This variant has higher
cell gene expression in tissue. ScRNA seq has necessary steps like RNA transmission rates. There was an average reduction in the neutralization
extraction, transcription into first-strand cDNA, 2nd-strand synthesis, efficiency of sera either from convalescent cases or from vaccines (Shen
and cDNA amplification (Haque et al., 2017). et al., 2021).
Moreover, via primary single-cell extraction, scRNA-seq should
conquer the effect of varied cell manifestation in tissues and attain Beta variant (B.1.351)
several evaluations with cell type identification, cell hierarchy recon It was first detected in October 2020 in South Africa. The mutation
struction, cell trajectory inference, and different capabilities (Hwang Number is 9. Defined in S proteins are D215G, E484K, D80A, R246I,
et al., 2018). Hence, the Next Generation Sequencing era will allow L18F, N501Y, D614G, K417N A701V. This variant has higher Re-
researchers to utilize this advanced technology to gain insights into infection rates. A high reduction in the sera’s neutralization efficiency
relative gene expression patterns and identify specific cell types asso has been seen in vaccines or convalescent persons (Zhou et al., 2021).
ciated with various diseases.
Gamma variant (P.1)
New variants of COVID-19 It was first detected in January 2021 in Brazil and Japan. The mu
tation Number is 12. Defined in S proteins are L18F, T20N, P26S, D138Y,
Emerging COVID-19 variants, characterized by genetic mutations, E484K, K417T, D614G, N501Y, V1176F, T1027I, R190S, H655Y (Faria
4
et al., 2021; Chavda and Apostolopoulos, 2022). This variant has a nearly 0.99 (Using Artificial Intelligence to Detect COVID-19 and
higher rate of re-infection. High reduction ability is found in sera’s Community-acquired Pneumonia Based on Pulmonary CT, 2023). A
neutralization efficiency from convalescent persons or vaccines. start-up, Infer Vision, involves medical imaging using deep learning
systems, quickly diagnosing COVID-19 cases by identifying lung fea
Omicron Variant (B.1.1.529) tures. AI-based algorithms diagnose COVID-19 infections employing 16
First detected in November 2021 in South Africa. Mutation Number basic parameters obtained from blood profiles, which may be a prom
is 0.30. Defined in S proteins are H655Y, N679K, P681H, R203K, G204R, ising alternative for RT-PCR testing in places with limited resources
E484, D614G, R93K, and G204R (Chavda and Apostolopoulos, 2022; (Soares et al., 2020).
Reynolds et al., 2022; , Chavda et al., xxxx). This variant has higher re- An AI system has been developed using a large computed tomogra
infection rates and greater transmission. Table 1 summarizes the impact phy (CT) database that assists in rapid NCP diagnosis and differentiates
of different COVID waves. it from other common pneumonia and normal controls (Zhang et al.,
2020; Wong et al., 2019; COVID-19 and artificial intelligence, 2023).
Role of AI for COVID 19 Block chain technology is a unique tool for the collection, validation,
verification, and confirmation of data and execution transactions (Ca
In COVID-19, AI is applicable in two ways. The former technique, sino et al., 2019). It is distinguished by its highly secure, patient-
like HealthMap, detects a new illness or epidemic. This necessitates the centered healthcare services, outstanding public health tracking,
use of algorithmic data. Secondly, a Pop-up symptom survey by Google outbreak management, and making decisions quickly and effectively.
and Facebook, data from Google search, and other sources can be used For monitoring and controlling the COVID-19 pandemic, both in
by AI to estimate the area needs for ventilators and intensive care units developed (to prevent overloading and stretching healthcare facility/
(Yu et al., 2021). Fig. 3 summarizes the pivotal role of AI in various laboratory infrastructure) as well as in developing, resource-constrained
genres for the termination of COVID-19. A web-based start-up, Bluedot, situations, a cost-effective block chain and AI-automated self-evaluator
employed ML to locate a cluster of pneumonia cases, forecast the and monitoring system have proved useful (The Emerging Role of
COVID-19 outbreak, and identify its geographic origin. In addition to Blockchain Technology Applications in Routine Disease Surveillance
this, AI can predict COVID-19 outbreaks by utilizing comparative and Systems to Strengthen Global Health Security - ProQuest, 2023). A study
multimodal data (Santosh, 2020). For detection, contact tracing, and looked into proposed or implemented block chain technologies for
effective tracking of COVID-19, smart devices can be used, including mitigating COVID-19 problems with the help of a detailed literature
mobile phones, cameras, watches, and many different wearable gadgets. survey. It was concluded that integrating blockchain technology with AI
AI can boost these mobile health applications (Maghdid et al., 2020). and geographic information systems (GIS) could improve the effective
Telemedicine can use programs such as AI4COVID-19 that employ 3-s ness and robustness of the public surveillance system (Elsevier, 2023;
coughing voice recording samples (Imran et al., 2020). The app can Sharma et al., 2022).
distinguish between COVID-19 cough and non-COVID-19 cough within Megvii, a start-up, performs fast screening of people from highly
2 min. populated areas and notices fever and high temperature, which may be
Zhang et al. reported that AI-based software could differentiate pa associated with COVID-19, through sophisticated face and body detec
tients with novel coronavirus pneumonia and other common pneumonia tion using infrared cameras (dual sensing), visible light, and thermal
based on thousands of CT scan images of human lungs and thus help to scanners. Alibaba Group AI-based applications use data parameters such
predict the chance of recovery or recurrence of COVID-19 (Zhang, et al., as self-reported health status, travel history, and contacts to identify
2020). COVID-19 detection neural network-COV-Net (Using Artificial COVID-19 cases (Pacheco Rocha et al., 2019). Srinivasa Rao and Vaz
Intelligence to Detect COVID-19 and Community-acquired Pneumonia quez defined an AI algorithm-based framework that would facilitate a
Based on Pulmonary CT, 2023), Multi-Objective Differential Evolution rapid recognition of infected cases considering novel coronavirus-
and convolutional neural networks (Singh and Kumar, 2020), COVID- related signs and symptoms through risk assessment and evaluation
Res-Net (Soares et al., 2020), and COVID MT-Net (Alom et al., 2020) (Srinivasa Rao and Vazquez, 2020). The required data can be collected
are different deep learning models used in the diagnosis and differen by a web- or smartphone-based survey. Alerts from the algorithm’s
tiation of novel coronavirus pneumonia (NCP) and other common interpretation can be sent to hospitals or mobile health units for health
pneumonia. COV-Net is a diagnostic method to recognize X-rays based checkups and case confirmations. Big Data and AI have potential ap
on machine learning. Experiment findings on two publicly available plications in COVID-19 management to monitor the spreading or
COVID-19 radiography databases revealed the feasibility of the pro transmission of viruses in real-time and plan and carry out public health
posed COV-Net in accurate COVID-19 recognition, with an accuracy of strategies as per those plans. It also monitors the effectiveness of
Table 1
SARS-CoV-2 Variants in three waves of the COVID-19 Pandemic (Chavda et al., 2022; El-Shabasy et al., 2022; Chavda and Apostolopoulos, 2022; Chavda et al., 2022;
Chavda and Apostolopoulos, 2022).
Subject First wave Second wave Third wave
Time 3rd month 2020 7th Month 2020 After December 2021
Preventive Over the top Mandatory Avoided by people
measure
Signs and Myalgia, cough, sore throat, fever, confusion, Pneumonia, communal cold, fever, breathlessness, Sudden death, Neurological effects, flu-like
indications breathlessness, headache, ageusia anosmia difficulties in respiration, and cough symptoms
Mortality Dreadful On higher side Lesser compared to previous
waves
Economy Negatively affected worldwide Affected Lesser affected than
previous waves
Social activities Stopped Reduced Remain without any change
Vaccination Not available In the process of examination Available
Re-infection No Yes Yes
Mutation No No Yes
Virus variant Epsilon, Eta, Iota, Kappa, Zeta, and Lambda Alpha, Beta, Gamma, and Delta Omicron
D614G SARS-CoV-2
5
Fig. 3. Role of AI in COVID-19 (created with Biorender.com).
interventions, repurposing the old drug molecules and finding new alanine aminotransferase (ALT) and aspartate aminotransferase (AST)
drugs, discovering potential vaccine molecules, and improving the are reversible after discontinuation of remdesivir per studies (Gupta
community’s responses to the ongoing pandemic (Srinivasa Rao and et al., 2020). Remdesivir use in patients with COVID-19 has been linked
Vazquez, 2020). to bradycardia, signs of worsening QT interval, and heart block, espe
Baricitinib is a biological molecule (Janus kinase inhibitor) used to cially in those with a history of cardiac conduction dysfunction. It is only
treat adult rheumatoid arthritis. AI quickly identified baricitinib’s administered through IV during hospitalization (Selvaraj et al., 2021).
antiviral activity (Zhou et al., 2020). Another in silico medicine approach Drug development is a complex process that involves choosing drug
has reported six novel molecules that may inhibit viral replication targets based on several factors, such as the disease’s mechanisms,
(COVID-19 and artificial intelligence, 2023). Zhang and coauthors existing knowledge, and potential risks and benefits (Vora et al., 2023).
employed network pharmacology analysis for systematic screening of In the case of COVID-19, the focus has been on targeting viral proteins to
Chinese natural constituents and identified 13 constituents beneficial in inhibit the virus’s replication and spread within the host. ACE2 is an
anti-COVID-19 special effects through the immune system modulation, essential protein in the human body that regulates blood pressure, and
viral replication regulation, and regulation of hypoxia cascade and in altering its function could have unintended and adverse effects (Shirb
flammatory pathways (Zhang et al., 2020). hate et al., 2021). Scientists may target established drug targets because
of their existing knowledge, proven safety profiles, and the need for
Drug repurposing rapid development of treatments during a pandemic. The lack of data for
unintended use of ACE 2 inhibitors restricted its use for COVID-19
As no particular therapy is available for treating mild-to-moderate (Angeli et al., 2022). However, it allows researchers to explore newer
COVID-19, researchers should investigate current drugs considering paths for that research.
their therapeutic potential, safety, acquaintance, and cost aspects The relationship between drugs (drug targets) and diseases [molec
(Cadegiani, 2020). Drug repurposing is a strategy that involves identi ular (protein) determinants in disease components] can be lighted by
fying already approved or investigational drugs, including those that novel approaches, such as network-based drug-disease proximity. It can
may have failed in clinical trials, and repurposing them for the treatment be a valuable tool for rapidly screening new indications for approved
of a disease or condition that they were not initially approved for. The drugs with well-established pharmacokinetics/pharmacodynamics,
large number of compounds requiring evaluation in preclinical trials safety, tolerability profiles, or previously unidentified adverse events
and the considerable proportion of randomized controlled trials (RCTs) (Cheng et al., 2018). Zhou et al. conducted a study that utilized a
that fail or raise concerns about toxicity are the primary reasons for the pharmacology-based network medicine platform to implement a drug
high costs and lengthy timelines associated with drug development. As a repurposing approach (Batra et al., 2020). Using network proximity
result, drug repurposing has emerged as a valuable strategy due to its analyses of drug targets and HCoV–host interactions in the human
potential to expedite timelines and reduce overall costs. Drug repur interactome, 16 potential anti-HCoV repurposable drugs were identi
posing is a promising technique that can effectively identify existing fied. The Z score was calculated using sequential network analysis, drug
drugs with potential therapeutic benefits for COVID-19 treatment. signatures, and HCoV-induced transcriptomic data in a human cell lines
Along with vaccines, different drugs such as remdesivir, molnupir study (Table 2). In the next part of the study, three promising drug
avir, and paxlovid were potentially aid in fighting the COVID-19 combinations (e.g., sirolimus in combination with dactinomycin,
pandemic. Elevated hepatic enzymes, anemia, diarrhea, rash, hypoten mercaptopurine with melatonin, and toremifene combined with
sion, and renal impairment are the side effects of remdesivir. Elevated emodin) were identified. The identified combinations had a unique
6
Table 2 mainly focused on (Cava et al., 2020). Wang et al. reviewed COVID-19
Potential repurposing drugs for COVID-19 (Batra et al., 2020). drug repurposing research, which utilizes clinical trials, computational
Drug name Z Known indication Potential for COVID-19 research, and in vitro protein binding experiments (Wang and Guan,
score treatment 2021). SARS-CoV-2 caused lung injury through the inhibition of ACE2.
Irbesartan − 5.98 Antihypertensive Na + -dependent Drug repurposing was carried out using an expression-based algorithm.
taurocholate co-transporting COL-3 (a chemically modified tetracycline) and CGP-60474 (a cyclin-
polypeptide (NTCP) inhibitor dependent kinase inhibitor) were promising repurposed drugs. These
(Wang, 2015) drugs target the genes causing lung injury (He and Garmire, 2020).
Toremifene − 3.23 Anticancer Selective estrogen receptor
MolGNN, a newly developed deep learning model, utilizes a graph
modulators (Zhou et al., neural network to identify chemical molecules computationally. This
2020) model, known for its powerful few-shot learning technique, demon
Camphor − 2.64 Antibiotic Selective estrogen receptor strates resilience when faced with limited training data. Through
modulators (Zhou et al.,
MolGNN, several multi-targeted drugs were predicted, targeting both
2020)
Equilin − 2.52 Hormone Interleukin-1 (IL-1), tumor human Janus kinases and the major protease of SARS-CoV-2. These
necrosis factor-α (TNF-α) targets are highly desirable for developing treatments that alleviate
suppressant and nuclear cytokine storm COVID-19 symptoms and restrict viral replication.
factor-κB (NF-κB) activation Notably, the compounds identified by MolGNN successfully reduced
inhibitor (Kim et al., 2013)
Mesalazine − 2.44 NSAID Papain-like protease
SARS-CoV-2-induced cell death (Liu et al., 2021).
inhibitor (Cheng et al., 2015) In another research, a graph-based study was carried out on 6 million
Mercaptopurine − 2.44 Anticancer Interleukin (IL)-10, tumor small molecules using a CAS-based biomedical graph using Neo4j
necrosis factor (TNF)-α, and (Version 4.1.0) software. STRING DB (version 11) was used to identify
CCL-2 inhibitor (Lee and
protein–protein interactions with nodes and networks. Out of 6 million,
Giuliani, 2019; Hojyo et al.,
2020) the top 50 drugs were identified through various sequential step ana
Paroxetine − 2.42 SSRI DNA topoisomerase 2-alpha lyses that can be repurposed and used for COVID-19. Of these, 11 drugs
(TOP2A) and DNA were under clinical trial for COVID-19 (listed in Table 3) (Knowledge
topoisomerase 2-beta Graph-Based Approaches to Drug Repurposing for COVID-19, 2023).
(TOP2B) inhibitor
(Zhou et al., 2020)
Sirolimus − 2.35 Immunosuppressants Interleukin 6 (IL-6)
suppressant (Skayem and Diagnosis and application of AI in COVID-19
Ayoub, 2020)
Carvedilol − 1.94 Beta-blocker Tubulin ligand complex and Proper diagnosis of COVID-19-infected patients is needed to target
IL-6 and TNF-α suppressant
effective treatment and transmission control. During COVID-19, severe
(Schlesinger et al., 2020)
Colchicine − 1.92 NSAID DNA topoisomerase 2-alpha and non-severe patients were distinguished by certain clinical features
(TOP2A) and DNA and not by systematic characterization of the complicated immunolog
topoisomerase 2-beta ical, physiological, and pathological response to the disease (Chavda and
(TOP2B) inhibitor Apostolopoulos, 2022; Chavda et al., 2022). For accurate detection,
Dactinomycin − 1.88 Antineoplastic Inhibitor of RNA virus
several sounds, such as groaning, breathing, chewing, and heartbeats,
replication, with known anti-
prion activity, potent Ebola can be integrated with a Natural Language Processing (NLP) model
virus inhibitor (Pineda et al., (Heidari et al., 2020). Intel® AI Builders member Huiying Medical has
2021) developed a medical imaging diagnostic answer that uses CT chest scans
Melatonin − 1.72 Hormone Ability to bind to viral DNA
to help with the early detection of coronavirus infections that comple
and RNA; inhibition of viral
ensemble; reduces SARS- ment routine science laboratory testing. CT scans and MRIs have played
CoV-2 virus replication and a pivotal role at the initial level of SARS-CoV-2 infected symptomatic
cytotoxicity (Salas Rojas patients (Malik et al., 2021). Hemdan et al. designed the COVIDX-Net
et al., 2021) network, which analyzes CXR images and ultimately aids radiologists
Quinacrine − 1.62 Antimalarial Blocks the SARS coronavirus
in decision-making. The dataset under consideration consisted of 50 X-
spike protein and
angiotensin-converting ray scans, and the network was trained using seven differentiated deep
enzyme 2 interaction CNN models of 25 verified COVID-19 instances. The model had an ac
(Ho et al., 2007) curacy of more than 80% (Hemdan et al., 2020). Heidari et al. developed
Emodin − 1.54 Anthraquinones Na + -dependent
VGG16, a CNN-based classifier to predict disease severity. Data from
taurocholate co-transporting
polypeptide (NTCP) inhibitor
approximately 8000 cases were used, with the outcomes correctly pre
(Wang, 2015) dicting COVID-19, COVID-19-infected pneumonia, and non-COVID-19
Oxymetholone − 1.53 Anabolic steroid Selective estrogen receptor
modulators (Zhou et al.,
Table 3
2020)
Potential drug for COVID-19 with ongoing clinical trials (Pirolli et al., 2021).
Drug Name Current Therapeutic Use
pattern referred to as the “Complementary Exposure” pattern. In this
Imatinib Kinase inhibitors
pattern, the drug targets were observed to impact the HCoV-host sub Tofacitinib Kinase inhibitors
network, while distinct neighborhoods within the human interactome Epigallocatechin gallate DNA metabolism-related inhibitors
network remained unaffected by the drugs (Zhou et al., 2020). Niclosamide Anthelmintics
In another study, 36 potential drugs were identified for COVID-19 Pictilisib Kinase inhibitors
Amiodarone Arrhythmia
treatment. Angiotensin-converting enzyme 2 (ACE2), the primary cell Leflunomide Rheumatoid Arthritis
receptor and key access of SARS-CoV-2 for entry into the cell to produce Colchicine Microtubule-regulating agents
the infection, was considered a target for drug repurposing. A pro Primaquine Antimalarial
tein–protein interaction network containing genes co-expressed with Ibrutinib Kinase inhibitors
Genistein Kinase inhibitors
ACE2 was constructed, and the genes associated with those drugs were
7
cases. The diaphragm region was removed during the preprocessing of contracting the coronavirus due to recent contact within a transmission
the chest X-rays. The original photos were preprocessed with a chain (Shahroz et al., 2021; Lee and Kim, 2021). Implementing
histogram-equalization technique followed by a bilateral filter. The CNN advancing technology into COVID-19 contact tracing is viewed as a
model was deployed after preprocessing. This study employed a transfer feasible approach that authorities, health practitioners, and IT techni
learning approach, achieving an impressive accuracy of 98.1% and cians should actively examine in limiting coronavirus spread (Mbunge,
sensitivity of 98.4%. The findings indicate that image preprocessing, 2020).
specifically generating pseudo-color images, can be utilized to develop a
deep learning computer-aided diagnosis (CAD) system for pneumonia Identifying, Tracking, and prediction of outbreak
diagnosis. This system holds the potential to provide accurate and reli
able results in the identification of pneumonia cases (Heidari et al., By leveraging AI, it becomes possible to locate clusters and “hot
2020). spots” where the virus is prevalent and analyze the infection level. This
Telemedicine, also called mobile health (mHealth), has emerged as a information enables successful contact tracing for individuals and allows
valuable tool for continuously monitoring and diagnosing COVID-19- for close monitoring of their status. AI can also contribute to predicting
infected patients, thereby reducing disease transmission (Haleem the progression of the infection and assessing the likelihood of its
et al., 2021). Additionally, the development of medical devices incor recurrence. Through these capabilities, AI plays a crucial role in
porating medical engineering and detection chemistry holds excellent enhancing surveillance and response efforts, aiding in effectively man
potential for effectively detecting and diagnosing the specific viral aging the pandemic (Vaishya et al., 2020). AI technology can track and
infection causing COVID-19. Such advancements can lead to improved predict the characteristics of the virus by analyzing infection trans
and efficient identification of active infections. Some of the recently mission risk data from social media platforms. It enables predicting the
developed devices are listed in the following Table 4. number of positive cases and fatalities in specific areas. AI also aids in
To enable continuous monitoring of health indicators and COVID-19- identifying vulnerable areas and populations, facilitating effective con
related symptoms, wearable devices such as sensors and patches have trol of virus transmission, and informing necessary treatment actions
been developed. These devices are worn close to the body and can detect (Abdulkareem and Petersen, 2021). The COUNTERACT (Recognizing
subtle changes in temperature, heart rate, oxygen saturation levels, and suspects and predicting the spread of Contagion based on Mobile Phone
other abnormal signs indicating active COVID-19 symptoms. Addition Location Data) system utilizes mobile phone location data to identify
ally, researchers have begun implementing mobile application-based potentially infectious places visited by individuals affected by a conta
monitoring systems using devices like the Fitbit band and Apple gious disease. The system calculates the incubation period for the
Watch. This progress highlights the importance of establishing a infected individual and then analyzes the locations they have visited
comprehensive population health database, leveraging predictive algo during that period using backpropagation. It is worth noting that a
rithms and machine learning techniques. Such databases can serve as the Samsung cell phone’s computational time was 15–18 times longer
foundation for developing mHealth platforms that provide guidance and compared to a dedicated HP edge line server setup (Ghayvat et al.,
support during pandemics like COVID-19. These platforms aim to utilize 2021). A systematic literature review (SLR) was done in another study to
the power of technology and data-driven approaches to enhance public get an overview of technological advances for detecting, monitoring,
health monitoring and response (Lionta et al., 2014). diagnosing, screening, surveillance, mapping hotspots, tracking, and
raising awareness to prevent and combat COVID-19 (Mbunge et al.,
Contact tracing 2021). In a study conducted by Vinay et al., a long short-term memory
(LSTM) network-based model was employed to predict the transmission
The rapid spread of COVID-19 worldwide has necessitated the of COVID-19. The study’s results revealed that the model achieved a
implementation of digital contact tracing methods using mobile appli high accuracy level of approximately 90–92%. This indicates the LSTM
cations and various technologies such as GPS, global contact details, network-based model’s effectiveness in predicting the virus’s spread
Bluetooth, card transaction data, social graph analysis, network-based (Chimmula and Zhang, 2020).
APIs, mobile tracking data, and physical system addresses (Shahroz
et al., 2021). Digital contact tracing processes, performed in real-time Epidemic prediction
using these applications, can be faster than non-digital methods. These
digital applications utilize AI and ML tools to analyze personal data, Epidemic prediction was one of the dominant problems in fore
enabling the identification of individuals who may be at risk of casting the progression of the disease. Thus, the challenge is to develop a
Table 4
Newly Developed Devices for COVID-19.
Name of AI Tool/ Composition Working Mechanism Use Reference
Technology
Ultrasensitive field- Capture antibodies bound to graphene This device can detect spike protein until a limit of Label-free detection of SARS- (Park et al.,
effect transistor-based sheets of the biosensor using a 1-pyrene- detection of 242 copies/mL in clinical samples CoV-2 S protein 2022)
biosensor butyric acid (PBA) N-hydroxy collected from the nasopharyngeal area. The signal
succinimide response is observed immediately upon antigen binding
ester linker and is stable over quantitative detection over a minute.
Electrochemical sensor Functionalized multiwalled carbon Detects reactive oxygen For real-time COVID-19 (Luong et al.,
for saliva/sputum nanotubes species (ROS) levels in sputum samples diagnosis and to monitor 2021)
lung health during infection
RapidPlex Immunosensor array based on four 1- Amperometric detection of SARS-CoV-2 N Simultaneous detection of (Torrente-
pyrene-butyric acid protein, anti-S1 IgG, anti-S1 IgM, and proteins. The SARS-CoV-2 Rodríguez
(PBA)-coated laser-engraved graphene data is wirelessly transferred to a user interface through antigen, antibodies, and et al., 2020)
working electrodes Bluetooth for remote reporting and monitoring. proteins
Breathalyzer system The sensor array of gold nanoparticles Alters the electric resistance, causing contraction and Identification of COVID-19 (Shan et al.,
functionalized with organic ligands enlargement of the nanomaterial film, which is signatures 2020)
triggered by chemical reactions when subjected to
exhaled air consisting of respiratory gases, volatile
organic compounds, and water vapor
8
predictive model to facilitate strategic planning in the healthcare system immune response, is possible using recombinant DNA (rDNA) and
to reduce the mortality rate (Desai et al., 2019). Hameni et al. developed sequencing methods. The rDNA-based vaccines are intended to be safer,
and evaluated the effectiveness of a deep learning model to predict the more effective, and low-cost when compared with conventional vac
pandemic. The normalized forward sensitivity index of the fundamental cines. The target epitopes may be included in combinations of specific
reproduction number, R0, was near 3, indicating that COVID-19 would epitopes or chimeric protein arrangements, which may provide a real
disappear without vaccines. In contrast, the number of new COVID-19 istic design capable of triggering beneficial humoral or cellular immune
cases would increase. However, the results were based on a short-term reactions. Molecular simulations, immune informatics, immunoge
study in a limited population (Desai et al., 2019). In another study nomics, structural biology, and bioinformatics techniques can all be
conducted by Zohair et al., a machine-learning approach was used to utilized to create advanced vaccine formulations in a rapid and precise
predict the mortality rate of COVID-19. The model showed an associa manner. Tables 5 and 6 summarize various applications of AI/ML in
tion between weather conditions and mortality rate. Higher temperature COVID-19 management.
and humidity showed a decline in the infection rate (Nkwayep et al., AI has significantly contributed to the expeditious advancement of
2020). medicinal treatments and vaccinations for the COVID-19 pandemic. By
examining extensive datasets, artificial intelligence (AI) models have
Protein structure prediction successfully identified promising medication candidates, made pre
dictions regarding protein structures, and optimized vaccine designs
AI can assist in significant protein structure prediction essential for (Chavda and Redwan, 2022; Khadela et al., 2023; Chavda et al., 2023).
virus entrance and replication. There are two strategies for predicting In addition, using artificial intelligence has expedited the process of
protein structure: template modeling and template-free modeling. The clinical trial simulations and enhanced patient monitoring, facilitating
former predicts protein structure using similar proteins as a template faster outcomes and ensuring the safety of participants (Sarmadi et al.,
sequence, whereas the latter predicts structure for proteins with un 2023). The integration of technology facilitated the optimization of the
known related structures (Swayamsiddha et al., 2021). This offers normally protracted pharmaceutical development process, leading to
insightful data, assisting rapid medication development (Bali and Bali, prompt interventions that mitigated the impact of the pandemic
2022). To predict the protein structures of membrane protein, protein (Table 6).
3a, nsp2, nsp4, and nsp6, as well as the papain-like C-terminal domain of
SARS-CoV-2, the AlphaFold algorithm of Google DeepMind used deep Virtual screening of therapeutic agents
residual networks (DRN) known as ResNets. The complex protein
structure of SARS-CoV-2 was determined using the deep convolutional SARS-CoV-2 has been sequenced, and continuous research aims to
neural network application, DeepTracer from high-resolution Cryo- understand its association with human cells. Theoretical Knowledge and
electron microscopy density maps and amino acid sequences (Vaishya AI are also used to screen small biomolecules with budding capability
et al., 2020). From an ultrastructure aspect, Google’s pioneering AI against COVID-19 using ML techniques (Soria-Guerra et al., 2015).
subsidiary, DeepMind, provided an AI prediction of the virus’s protein Studies have indicated that the interaction between coronavirus spike
structure based on genome sequencing. This has been utilized as a protein (S-protein) and specific human host receptors, such as the ACE-2
foundation for additional AI protein folding experiments to produce receptor, is responsible for viral infection. Theoretically, compounds
possible medications that can influence viral replication or function, that can hinder this host-viral interaction can potentially weaken the
with several future therapeutics currently under investigation (Elwazir binding between the S-protein and ACE-2 receptors (Chavda and
and Hosny, 2020). Apostolopoulos, 2023), thereby impeding viral entry into host cells
(Fig. 4). Virtual screening has emerged as a crucial tool in searching for
AI in the development of therapeutics and vaccine therapies targeting COVID-19. By employing computational methodol
ogies, researchers efficiently evaluated extensive collections of chem
Molecular dynamics and molecular docking are used to identify icals to discover possible inhibitors of the essential proteins of the virus.
drugs of synthetic and natural origin to target the SARS-CoV-2 protein By simulating chemical interactions, researchers could identify medi
(Chavda et al., 2023). These computational methods assist in scruti cation candidates that show promise regarding their prospective anti
nizing synthetic or natural origin drugs with the binding capacity to viral characteristics. This expedited methodology facilitated the
active sites of the targets (Pinzi and Rastelli, 2019). In molecular discovery of prospective repurposed pharmaceuticals and innovative
docking, mathematical models are utilized to calculate the binding free compounds, optimizing the progression towards in-vitro and clinical
energy score with better stability of the complex. Molecular dynamics assessments targeting the novel coronavirus. It is essential to note that
simulations can determine the microscopic interactions between drug any intervention, whether it targets ACE2 or any other protein or
molecules and 3D structures (Nair and Miners, 2014). AI holds signifi pathway, comes with risks. The key is to weigh those risks against the
cant promise in exploring the vast chemical space and accelerating the benefits and to continuously monitor and adjust strategies based on
development of novel therapeutics. accumulating knowledge and evidence.
In contrast, the conventional vaccine development process, which There are two possibilities for such interaction. 1) S-Protein of the
involves using attenuated or killed agents, can take up to 15 years. This virus at its host receptors region of host hindering viral recognition of
lengthy process entails scaling up the desired microorganism, ensuring the host cells or 2) S-Protein: human ACE2 interface debilitating the
ideal growth conditions, effectively inactivating it, and evaluating the host-viral interaction. Currently, 77 ligands have been effectively
immunogenicity of the new vaccine. However, with advancements such identified, of which 24 got regulatory approval from the FDA and similar
as understanding critical epitopes responsible for immune reactions, regulatory agencies that fulfilled any of these criteria (Fig. 5) (Sharma,
rDNA technology, the genetic information of viruses, and bioinformatics 2021; Batra et al., 2020). Although the enormous chemical space (mil
tools’ upgradations, it is now possible to develop new vaccines within a lions to billions of biomolecules) can be conceivably investigated, nearly
much shorter timeframe of 1–2 years. AI and these scientific advance 9000 applicant compounds were considered inferable from the
ments are poised to revolutionize the speed and efficiency of vaccine tremendous computational cost of group docking studies used in their
development (Soria-Guerra et al., 2015). In silico predictions help to methodology (Huang et al., 2004; Ergin et al., 2022).
define targets and are advantageous for the rapid development and It has been highlighted in Fig. 6, an overall work process that can be
commercial availability of COVID-19 vaccines in epidemic situations utilized for the dynamic screening of molecules with a target binding
(Chavda and Apostolopoulos, 2022). energy. Their binding to S-Protein: human ACE2 receptor interface is
Identification of isolated epitopes, which can stimulate a specific used by this working method to screen therapeutic molecules
9
Table 5
Various Applications of AI in COVID-19 Management.
Applications Software/ Owner How AI works References
Algorithms
Discovery of New CogMol Bluedot Innovation, Toronto, Target-Specific and Selective Drug Design for COVID-19 (Chenthamarakshan et al., 2020)
therapeutic agents Canada Using Deep Generative Models
against SARS-CoV-2 De Novo John Brownstein, Boston, USA Designing of chemical molecules and determination (Srinivasan et al., 2021)
Molecular Design of Vina scores, which are considered as a measure of
Monte Carlo Tree binding to S-
Search protein at the host receptor region or to the S-protein/ACE
Recurrent neural 2 receptor interface
networks
Protein Structure AlphaFold Google Deepmind, London, Deep residual networks (DRN) known as ResNets are used (Lee and Kim, 2021; Vaishya
Prediction algorithm UK to predict the protein membrane structure and membrane et al., 2020)
components, such as the C terminal, nsp6, and nsp4, which
can provide a massive advantage for discovering the drug
substances.
Development AutoDock Vina The Scripps Research Useful in screening (by molecular docking) promising (Yu et al., 2020)
of Institute., California, USA drugs to target the structural and non-structural protein
Therapeutics sites of the new coronavirus
Bioinformatics Tools for Multiscale For visualization of viral coats (large-scale molecular (Waman et al., 2021)
Molecular Structure models structure)
Recognition ViewDock To view screen docked ligand orientation
Volume viewer To visualize density maps
Multalign viewer To display sequence alignments with crosstalk to any
associated structure
In Early Diagnosis COVNet: Wang and Wong, Waterloo AI It is a deep learning-based algorithm used to differentiate (Kedia et al., 2021)
COVID-19 Institute, Waterloo, Canada between community-acquired pneumonia (CAP) and
detection neural COVID-19 based on 2D and 3D visuals seen in the
network volumetric CT scan of the chest.
Detection and annotation VirHostNet Elixir bio.tools, the European Understanding molecular mechanisms underlying virus (Mbunge et al., 2021)
Virus Bioinformatics Center replication and pathogenesis
and FAIRsharing.org
CORDITE CORona Drug InTEractions Carrying out meta-analyses on potential drugs and
database identifying potential drug candidates for clinical trials
University of Marburg
CoVex University of Hamburg Identifying already approved drugs that could be
Germany repurposed to treat COVID-19
P-HIPSTer Columbia University An algorithm that exploits both sequence-based and
Department of Systems structure-based information to infer interactions between
Biology pathogen and human proteins.
In reducing the workload Chatbox- Clara HCL Tech AI-driven technologies could reduce the workload of the (Mbunge et al., 2021; Chimmula
of healthcare medical and clinical healthcare staff by the digitization of and Zhang, 2020; Nkwayep
professionals the records of patients, and professional training can et al., 2020)
reduce the contact of these professionals with the patients.
Also, personalized treatment options and telemedicine can
be made available, which helps to avoid exposure to the
virus.
Contact tracing of A14-COVID-19 Apple and Google Partnered AI-enabled preliminary diagnosis for COVID-19 from (Imran et al., 2020)
individual Programmed, USA cough samples
Development of Vaccine Vaxign-ML University of Michigan, USA An ML-based algorithm helps identify the suitable vaccine (Ong et al., 2020)
candidate for further evaluation.
(Lalmuanawma et al., 2020). In particular, the screening uses auto- been shown, which makes the SARS-CoV-2 theory protease an attractive
docking or subatomic display information set to prepare and validate medication target, especially since its gem structure has been decided as
ML models. This extends the search space, and millions of possible of late. They extracted an estimated 606 million mixtures from the ZINC
therapeutic molecules are screened. knowledge repository. To begin, all mixtures in the three-dimensional
Virtual screening is a handy application to identify promising mol structure were checked for form closeness (Zhang et al., 2020) against
ecules as a starting point in drug discovery. As virtual screening becomes a pre-selected set of known and co-crystalized compounds. The rapid
increasingly essential and prominent in the pharmaceutical industry, screening of huge knowledge bases based on established covers as
this approach has witnessed a rapid surge in successfully identifying format takes into account such a coarse GPU-quickened convention. Past
potential therapeutic candidates (Fig. 7) (Pirolli et al., 2021). virtual screening activities were confined to more modest compound
Virtual Screening is a generally utilized strategy at the beginning of libraries as a whole.
medication revelation that permits the distinction of possibly bioactive
mixes at high throughput (Fig. 7) (Lionta et al., 2014). In a few previous The role of AI in clinical trial
undertakings, combining diverse virtual screening techniques led to the
disclosure of intense inhibitors (Lionta et al., 2014; Xu et al., 2020). It is Whenever a new disease or disorder arises, there is a need to discover
a promising way to deal with distinguishing potential medication rise a therapy that can efficiently mitigate that particular disorder. Drug
against the uniformly rising SARS-CoV-2 infection due to its rate and discovery is the process that can associate the functioning of a syn
expense, mainly when time is of quintessence. For HIV and HCV, the thetically derived molecule to the disease by its therapeutic action to
hindrance of proteases as a treatment against viral contamination has eradicate it from the human body. Preclinical and clinical trial study
10
Table 6
Application of different models in Detecting, Diagnosing, and Predicting COVID-19.
Application Model Key role Purpose of model References
Name
Machine Support vector machine (SVM) A most powerful tool for solving regression problems Detection of X-ray images,
Learning and classification Detection of Blood and (Guhathakurata
(ML) detection, prediction, and forecasting of the COVID- Urine Test, et al., 2021)
19 pandemic. Classification of CT Images,
Prediction of Clinical Data.
Random Forest (RF) Used to solve classification and regression problems Identification of Blood tests and (Khan et al., 2021)
Clinical Data.
Severity assessment of CT Images,
Prediction of Clinical data and CT
images.
Logistic regression (LR) When the target variable is categorical, LR is used as a Screening of CT Images, (Lalmuanawma
statistical tool and regression analysis that uses a Identification and Diagnosis of Times et al., 2020)
logistic function to model a binary dependent series,
variable. Prediction of RT-PCR Test, Clinical
Data, and CT Images.
Decision tree (DT) Used for solving regression and classification For detection, diagnosis, and (Shanbehzadeh
problems. prediction of Clinical Data et al., 2021)
Other ML Xgboost, (MLP) multilayer perceptron, – Detection, forecasting of COVID-19, (Heidari et al.,
methods k-means, neural networks, and prediction, screening. 2022)
Gaussian process regression (GPR)
Deep Learning Convolutional neural network (CNN) Medical Image Analysis COVID-19 screening, diagnostics, (Mei et al., 2020)
(DL) classification, prediction, and
forecasting of X-ray and CT images
Long short-term memory (LSTM) Recurrent neural network (RNN), which has the Widely used for the time series (Shorten et al.,
capability of learning order dependence in sequence sequential data prediction 2021)
prediction problems
Fig. 4. Interaction of S-Protein of virus and host receptors (created with Biorender.com).
data is required to approve any new therapeutic molecule. A consider The term “AI” was coined by McCarthy in the 1950 s (Corso et al.,
able amount of data is generated through these studies. AI and ML tools 2016), and it gained significant popularity in the 1970 s when it was
may assist in data management and make the research work less chal employed in MYCIN to provide diagnostic decision support. The drug
lenging (Zhang and Lu, 2021). development process typically spans 10 to 15 years, with a significant
11
Fig. 5. Flow sheet of the ML-based approach to predict FDA-approved drugs for COVID-19 (created with Biorender.com).
Fig. 6. Instance of AI applications at different stages of COVID-19 (created with Biorender.com).
portion of that time dedicated to clinical trials, which also entail sub decoder. This model appropriately used the latent space to design new
stantial financial investments. However, using AI can reduce the cost drug molecules with optimized and desired therapeutic properties (Liu
and time required for drug development. et al., 2018). A Bayesian machine-learning algorithm called BANDIT was
AI and ML in clinical development can eliminate chemical moieties proposed to predict drug-binding targets by integrating various data
that are less effective from the beginning, ultimately saving high sources. A novel therapeutic repositioning opportunity is suggested by
research costs and time consumption. Chances to achieve desired results BANDIT, which also explains clinical data that were previously unrec
can be increased using targeted designing tools that accurately provide ognized. BANDIT is an effective and precise platform that can hasten
outcomes. AI can reduce conventional drug design methods and risk medication discovery and clinical setting implementation (A Bayesian
analysis challenges and obstacles. Delivering therapeutic agents with machine learning approach for drug target identification using diverse
high activity and less danger has become a challenge. AI comes into data types, 2023).
action to diminish these challenges. AI and ML give algorithms that can The Trials.ai platform uses artificial intelligence via NLP (Natural
aid in the synthesis of molecule, protein, and peptide interaction, Language Processing) and ML to accelerate and streamline clinical trial
structure elucidation, prediction of toxicity, bioactivity prediction by methods. This platform also assists in recognizing possible hazards and
physicochemical properties, disease diagnostics, monitoring the release obstructions to a successful trial completion (such as inclusion/exclu
of drugs, quantitative structure activity-relationship, design of experi sion criteria or outcome assessments) through uploaded protocol data
ments, and data regulation and management are available. and NLP (Trials.ai – Our Smart Protocol technology helps teams design
For expressing interactions between new chemical entities, graphs better clinical trials with AI [Internet]. [cited, 2023). The SPIRIT-AI and
are a standard data format. A variational auto-encoder model was pro CONSORT-AI extensions were created parallel for clinical trial protocols
posed, which analyzes the graph-structured data using an encoder and and trial reports. The SPIRIT-AI and CONSORT-AI Steering Group were
12
Fig. 7. Potential Drug Discovery of SARS-CoV-2 by Structure-based Virtual Screening (created with Biorender.com).
established to monitor the conduct and method of the study. It is Ethical and regulatory concerns
comprised of 15 global specialists. When appropriate, it is utilized to
expand or clarify this guideline to help report clinical studies for AI The ethical and regulatory framework for AI in the healthcare sector
intervention (Reporting guidelines for clinical trial reports for in is multifaceted. Addressing concerns about patient privacy, data secu
terventions involving artificial intelligence: the CONSORT-AI extension rity, and biases inherent in AI systems is imperative. Patient privacy is
| Nature Medicine [Internet]. [cited, 2023). one of the cornerstones of healthcare ethics, and any technological
advancement, including AI, must uphold the sanctity of patients’ per
Challenges in the adaptation of AI sonal and medical information (Murdoch, 2021; Markus et al., 2021).
Ensuring that sensitive patient data remains confidential, with appro
In recent years, significant advancements in AI have led to trans priate safeguards in place, is of non-negotiable significance in this era of
formative effects across multiple sectors, including the healthcare in data-driven healthcare.
dustry. AI can augment clinical decision-making, optimize healthcare Data security in healthcare transcends the mere protection of infor
workflows, and increase patient outcomes. Nevertheless, the incorpo mation. It encompasses the development of robust cybersecurity mea
ration of AI inside the therapeutic environment presents notable ob sures, data encryption protocols, and secure data transmission, all of
stacles (Bajwa et al., 2021; Dash et al., 2019). This review article which are essential to shield healthcare data from breaches and cyber
examines the difficulties mentioned above via a scientific lens, offering a threats (Seh et al., 2020).
comprehensive analysis supported by references to contemporary Intrinsic biases in AI algorithms can inadvertently perpetuate dis
studies and advancements. parities and inequalities within healthcare. These biases can result from
historical imbalances in training datasets, among other factors (Norori
Data quality and availability et al., 2021). Using data preparation techniques, Buolamwini and Gebru
identified such biases in facial recognition systems (Raji and Buo
One of the primary obstacles encountered in implementing AI within lamwini, 2019). Recognizing and rectifying these biases in AI systems is
the healthcare sector is the adequacy and accessibility of data. AI al an ethical and regulatory imperative to ensure that healthcare AI fosters
gorithms largely depend on extensive datasets of considerable size and equitable and unbiased healthcare outcomes.
superior quality for training and validation. Within the clinical setting, The ethical implications encompass the potential of AI to influence
data can exhibit limitations such as incompleteness, noise, and healthcare inequities and amplify pre-existing inequalities. Balancing
authenticity. Besides, collecting data without consent is unethical and innovation with sound regulation stands as an ongoing challenge
profoundly challenges a patient’s right to privacy. In order to ensure the (Morley et al., 2022). The dynamism of AI technologies necessitates
robustness and generalizability of AI models, researchers must negotiate regulatory frameworks that are both robust and flexible. The European
these limits (Price and Cohen, 2019; Obermeyer and Emanuel, 2016). Commission’s White Paper on AI Regulation (2020) outlines regulatory
proposals rooted in ethical principles and can be considered a step for
ward to strengthen ethical principles. Regularly assessing AI systems for
Interpretability and explainability
ethical issues is critical. Research by Jobin et al. provides guidelines for
ethical audits and monitoring (Jobin et al., xxxx).
AI models, specifically deep learning neural networks, have highly
The resolution of ethical dilemmas in artificial intelligence necessi
complex calculations. Such functions are often opaque systems, and it is
tates adopting a comprehensive approach that encompasses scientific
challenging to comprehend the rationale behind their specific outcomes
inquiry, regulatory measures, and proactive design methodologies (Naik
(Linardatos et al., 2021). In clinical practice, the concepts of interpret
et al., 2022; Gerke et al., 2020). By integrating these concepts, it is
ability and explainability are the most critical pillars in establishing trust
possible to guarantee that AI technologies contribute to the betterment
and guaranteeing the safety of patients. So, researchers must pursue
of society while upholding ethical standards and human values.
developing methodologies that offer transparency, easy interpretability,
and logical reasoning of the AI-based decision-making processes
(Amann et al., 2020; Caruana et al., 2015).
13
Clinical validation and integration harness AI’s capabilities. This collaborative effort should be guided by a
commitment to upholding ethical standards and prioritizing the well-
For AI solutions to be adopted in clinical practice, they must undergo being of patients, ensuring that AI becomes a valuable and ethically
a rigorous validation process, including safety, effectiveness, and prac sound tool in healthcare.
ticality assessments (Tsopra et al., 2021). Healthcare systems are com
plex and multifaceted, with established workflows and electronic health
record (EHR) systems integral to clinical practices. Integrating these AI Harmonization
systems into existing healthcare infrastructure, workflows, and elec
tronic health records (EHRs) is a significant challenge. Additionally, AI Due to the rapidly evolving industry, adaptation of AI is challenging.
is data-dependent, and understanding how to harness these technologies International AI collaboration is essential to pool resources, expertise,
to their potential demands a learning curve. The training of personnel and knowledge. This global collaboration can harmonize rules and
for AI implementation is a crucial but complex task. Achieving seamless regulations governing AI in healthcare across different countries,
integration is essential to ensure that AI solutions are user-friendly and enabling a consistent and ethical approach to AI implementation (Bohr
do not disrupt clinical practices (Davenport and Kalakota, 2019). and Memarzadeh, 2020). Additionally, the successful integration of AI
in healthcare relies on robust collaboration between healthcare workers
Cost and resource constraints and policymakers (Hazarika, 2020). This synergy ensures that AI tech
nologies are not only ethically sound but also tailored to meet the real-
Integrating AI within the healthcare sector can incur substantial world needs of healthcare, ultimately benefiting both patients and the
expenses, encompassing infrastructural and human resource in broader healthcare ecosystem (Mittelstadt, 2019).
vestments. Smaller healthcare facilities may encounter difficulties while
allocating the requisite financial resources and human capital for the Conclusion
integration of AI. Crucially, personnel training for AI implementation,
an indispensable endeavor, is both time-intensive and economically The role of AI and ML in harnessing therapeutic and vaccine devel
demanding. Addressing this challenge requires a concerted effort to opment is immense and has fastened the development cycle to circum
ensure that AI technologies are accessible to a broad spectrum of vent unmet patient needs. From decoding the virus’s structure to
healthcare providers and patients (Norgeot et al., 2020). Additionally, identifying the efficacious molecules, AI is always used and proven to be
there is a need to integrate AI technologies into educational curricula, efficacious and cogent. The algorithm-based big data analysis will
facilitating the adaptation of new healthcare professionals to AI tools reduce therapeutic uncertainty. Digital Health strongly encourages re
and accelerating their on-field proficiency (Mir et al., 2023). searchers doing AI intervention-based clinical trials to follow the new
extension guidelines SPIRIT-AI and CONSORT-AI. As we are facing the
Continual learning and adaptation third outbreak of COVID-19, the usage of AI and ML will help us in ca
pacity building for preventing future outbreaks. There is a need to have a
The medical field is distinguished by its constant evolution, marked harmonized legal framework and ethical considerations for sharing the
by the emergence of novel research findings and evolving treatment data, which help us effectively manage medical interventions.
protocols. In this dynamic landscape, AI models must demonstrate the
ability to adapt and remain up-to-date in response to these ongoing
changes (Komorowski et al., 2018). For instance, the rapid mutation of Declaration of competing interest
the SARS-CoV-2 virus and the emergence of new strains of COVID-19
exemplify the dynamic landscape of medicine, where timely updates The authors declare that they have no known competing financial
to diagnostic and treatment approaches are crucial (Chavda et al., interests or personal relationships that could have appeared to influence
2022). the work reported in this paper.
The scientific challenge involves achieving a delicate balance be
tween continuous learning and adaptation while preserving model sta Data availability
bility, necessitating unceasing research endeavors. AI models for
COVID-19 diagnostics and treatment planning must quickly incorpo No data was used for the research described in the article.
rate information about new virus variants and adapt their recommen
dations accordingly (Comito and Pizzuti, 2022). This requires ongoing Acknowledgments
data collection, model refinement, and validation processes to ensure
that AI systems remain accurate and reliable in an ever-changing VPC wants to dedicate this article to the 75th-year celebration of L M
healthcare environment (Vora et al., 2023). College of Pharmacy.
Limited generalization References
The generalizability of AI models built in a specific therapeutic A Bayesian machine learning approach for drug target identification using diverse data
types | Nature Communications [Internet]. [cited 2023 Jul 10]. Available from:
scenario may be limited when applied to other settings. The broad https://www.nature.com/articles/s41467-019-12928-6.
applicability of AI solutions may be hindered by factors such as varia Abdulkareem M, Petersen SE. The Promise of AI in Detection, Diagnosis, and
tions in patient populations, healthcare practices, and data collection Epidemiology for Combating COVID-19: Beyond the Hype. Frontiers in Artificial
Intelligence [Internet]. 2021 [cited 2023 Jul 10];4. Available from: https://www.
methods (Jiang et al., 202). Academic researchers are working to frontiersin.org/articles/10.3389/frai.2021.652669.
enhance AI model transferability across diverse healthcare settings Aleem, A., Akbar Samad, A.B., Vaqar, S., 2023. Emerging Variants of SARS-CoV-2 and
(Zech et al., 2018). Novel Therapeutics Against Coronavirus (COVID-19). In: StatPearls [Internet].
Treasure Island (FL): StatPearls Publishing; 2023 [cited 2023 Jul 9]. Available from:
Integrating AI into clinical settings holds immense potential for
http://www.ncbi.nlm.nih.gov/books/NBK570580/.
transforming the healthcare industry. However, addressing the scientific Alom, M.Z., Rahman, M.M.S., Nasrin, M.S., Taha, T.M., Asari, V.K., 2020. COVID_MTNet:
challenges outlined in this article is essential to ensure AI’s secure and COVID-19 Detection with Multi-Task Deep Learning Approaches [Internet]. arXiv;
effective adoption in clinical practice (Davenport and Kalakota, 2019). 2020 [cited 2023 Jul 9]. Available from: http://arxiv.org/abs/2004.03747.
Amann, J., Blasimme, A., Vayena, E., Frey, D., Madai, V.I., 2020 Nov. Explainability for
Researchers, healthcare professionals, pharmaceutical industries, and artificial intelligence in healthcare: a multidisciplinary perspective. BMC Med. Inf.
policymakers must collaborate to overcome these obstacles and fully Decis. Making 30 (20), 310.
14
Angeli, F., Reboldi, G., Trapasso, M., Zappa, M., Spanevello, A., Verdecchia, P., 2022 Sep. Chavda, V.P., Patel, A.B., Vaghasiya, D.D., 2022. SARS-CoV-2 variants and vulnerability
COVID-19, vaccines and deficiency of ACE2 and other angiotensinases. Closing the at the global level. Journal of Medical Virology [Internet]. 2022 Mar;94:2986–3005.
loop on the “Spike effect”. Eur. J. Intern. Med. 103, 23–28. Available from: https://doi.org/10.1002/jmv.27717.
Arslan, J., Benke, K.K., 2021. Artificial Intelligence and Telehealth may Provide Early Chavda, V.P., Vuppu, S., Mishra, T., Kamaraj, S., Patel, A.B., Sharma, N., et al., 2022.
Warning of Epidemics. Frontiers in Artificial Intelligence [Internet]. 2021 [cited Recent review of COVID-19 management: diagnosis, treatment and vaccination.
2023 Jul 9];4. Available from: https://www.frontiersin.org/articles/10.3389/ Pharmacological Reports [Internet]. 2022;Online first-Online first. Available from:
frai.2021.556848. https://doi.org/10.1007/s43440-022-00425-5.
Artificial Intelligence and COVID-19: Deep Learning Approaches for Diagnosis and Chavda, V., Bezbaruah, R., Valu, D., Desai, S., Chauhan, N., Marwadi, S., et al., 2023.
Treatment | IEEE Journals & Magazine | IEEE Xplore [Internet]. [cited 2023 Jul 9]. Clinical Applications of “Omics” Technology as a Bioinformatic Tool. In:
Available from: https://ieeexplore.ieee.org/document/9115663. Bioinformatics Tools for Pharmaceutical Drug Product Development [Internet]. 2023
Artificial intelligence–enabled rapid diagnosis of patients with COVID-19 | Nature [cited 2023 May 22]. p. 117–45. Available from: https://doi.org/10.1002/
Medicine [Internet]. [cited 2023 Jul 9]. Available from: https://www.nature.com/ 9781119865728.ch7.
articles/s41591-020-0931-3. Chavda, V.P., Patel, K., Patel, S., Apostolopoulos ,V., 2023. Artificial Intelligence and
Bajwa, J., Munir, U., Nori, A., Williams, B., 2021 Jul. Artificial intelligence in healthcare: Machine Learning in Healthcare Sector. In: Bioinformatics Tools for Pharmaceutical
transforming the practice of medicine. Future Healthc J. 8 (2), e188–e194. Drug Product Development [Internet]. 2023 [cited 2023 May 22]. p. 285–314.
Bali, A., Bali, N., 2022. Chapter 8 - Role of artificial intelligence in fast-track drug Available from: https://doi.org/10.1002/9781119865728.ch13.
discovery and vaccine development for COVID-19. In: Chang V, Abdel-Basset M, Chavda, V.P., Hanuma Kumar Ghali, E.N., Yallapu MM, Apostolopoulos V. Therapeutics
Ramachandran M, Green NG, Wills G, editors. Novel AI and Data Science to tackle Omicron outbreak. Immunotherapy. :10.2217/imt-2022–0064.
Advancements for Sustainability in the Era of COVID-19 [Internet]. Academic Press; Chavda, V.P., Teli, D., Balar, P.C., Vaghela, D., Solanki, H.K., Vaishnav, A., et al., 2023.
2022 [cited 2023 Jul 10]. p. 201–29. Available from: https://www.sciencedirect. Potential Anti-SARS-CoV-2 Prodrugs Activated by Phosphorylation and Their Role in
com/science/article/pii/B9780323900546000064. the Aged Population. Molecules 28 (5).
Bali, A., Bali, N., 2022 [cited 2023 Oct 18].. Role of artificial intelligence in fast-track Chavda, V.P., Valu, D.D., Parikh, P.K., Tiwari, N., Chhipa, A.S., Shukla, S., et al., 2023.
drug discovery and vaccine development for COVID-19. In: Novel AI and Data Conventional and Novel Diagnostic Tools for the Diagnosis of Emerging SARS-CoV-2
Science Advancements for SustAinability in the Era of COVID-19 [internet]. Elsevier, Variants. Vaccines. 11 (2).
pp. 201–229. Available from: Chen, J., See, K.C., 2020 Oct 27. Artificial Intelligence for COVID-19: Rapid Review.
Batra, R., Chan, H., Kamath, G., Ramprasad, R., Cherukara, M.J., Sankaranarayanan, S.K. J. Med. Internet Res. 22 (10), e21476.
R.S., 2020 Sep 3. Screening of Therapeutic Agents for COVID-19 Using Machine Cheng, K.W., Cheng, S.C., Chen, W.Y., Lin, M.H., Chuang, S.J., Cheng, I.H., et al., 2015
Learning and Ensemble Docking Studies. J. Phys. Chem. Lett. 11 (17), 7058–7065. Mar. Thiopurine analogs and mycophenolic acid synergistically inhibit the papain-
Bhattacharya, S., Singh, A., Hossain, M.M., 2019 Sep 2. Strengthening public health like protease of Middle East respiratory syndrome coronavirus. Antiviral Res. 115,
surveillance through blockchain technology. AIMS Public Health. 6 (3), 326–333. 9–16.
Bohr, A., Memarzadeh, K., 2020. The rise of artificial intelligence in healthcare Cheng, F., Desai, R.J., Handy, D.E., Wang, R., Schneeweiss, S., Barabási, A.L., et al., 2018
applications. Artif. Intell. Healthcare. 25–60. Jul 12. Network-based approach to prediction and population-based validation of in
Cadegiani, F.A., 2020 Dec. Repurposing existing drugs for COVID-19: an endocrinology silico drug repurposing. Nat. Commun. 9 (1), 2691.
perspective. BMC Endocr. Disord. 20 (1), 149. Chenthamarakshan, V., Das, P., Hoffman, S.C., Strobelt, H., Padhi, I., Lim, K.W., et al.
Cardiovascular Implications of Fatal Outcomes of Patients With Coronavirus Disease CogMol: Target-Specific and Selective Drug Design for COVID-19 Using Deep
2019 (COVID-19) | Cardiology | JAMA Cardiology | JAMA Network [Internet]. Generative Models [Internet]. arXiv; 2020 [cited 2023 Jul 10]. Available from:
[cited 2023 Jul 9]. Available from: https://jamanetwork.com/journals/ http://arxiv.org/abs/2004.01215.
jamacardiology/fullarticle/2763845. Chimmula, V.K.R., Zhang, L., 2020 Jun. Time series forecasting of COVID-19
Caruana, R., Lou, Y., Gehrke, J., Koch, P., Sturm, M., Elhadad, N., 2015. Intelligible transmission in Canada using LSTM networks. Chaos Solitons Fractals 1 (135),
Models for HealthCare: Predicting Pneumonia Risk and Hospital 30-day 109864.
Readmission. In: Proceedings of the 21th ACM SIGKDD International Conference on Choi, J.Y., Smith, D.M., 2021 Nov. SARS-CoV-2 Variants of Concern. Yonsei Med. J. 62
Knowledge Discovery and Data Mining [Internet]. New York, NY, USA: Association (11), 961–968.
for Computing Machinery; 2015 [cited 2023 Oct 12]. p. 1721–30. (KDD ’15). Comito, C., Pizzuti, C., 2022 Jun. Artificial intelligence for forecasting and diagnosing
Available from: https://doi.org/10.1145/2783258.2788613. COVID-19 pandemic: A focused review. Artif. Intell. Med. 128, 102286.
Casino, F., Dasaklis, T.K., Patsakis, C., 2019 Mar. A systematic literature review of Corso, G., Alisi, M.A., Cazzolla, N., Coletta, I., Furlotti, G., Garofalo, B., et al., 2016 Sep.
blockchain-based applications: Current status, classification and open issues. A Novel Multi-step Virtual Screening for the Identification of Human and Mouse
Telematics Inform. 1 (36), 55–81. mPGES-1 Inhibitors. Mol Inform. 35 (8–9), 358–368.
Cava, C., Bertoli, G., Castiglioni, I., 2020 Apr 6. In Silico Discovery of Candidate Drugs COVID-19 and artificial intelligence: protecting health-care workers and curbing the
against Covid-19. Viruses 12 (4), 404. spread - The Lancet Digital Health [Internet]. [cited 2023 Jul 9]. Available from:
Chavda, V.P., Apostolopoulos, V., 2022. Omicron variant (B.1.1.529) of SARS-CoV-2: https://www.thelancet.com/journals/landig/article/PIIS2589-7500(20)30054-6/
Threat for the elderly? Maturitas. 2022 Apr;158:78–81. fulltext.
Chavda, V.P., Apostolopoulos, V., 2023. Mosaic receptor binding domain nanoparticles: Dash, S., Shakyawar, S.K., Sharma, M., Kaushik, S., 2019 Jun 19. Big data in healthcare:
towards fourth-generation vaccination. Nanomedicine [Internet]. 2023 Aug 1 [cited management, analysis and future prospects. J. Big Data. 6 (1), 54.
2023 Oct 28];18(19):1223–6. Available from: https://doi.org/10.2217/nnm-2022- Davenport, T., Kalakota, R., 2019 Jun. The potential for artificial intelligence in
0316. healthcare. Future Healthc J. 6 (2), 94–98.
Chavda, V.P., Apostolopoulos, V., 2022 May. Global impact of delta plus variant and Desai, A.N., Kraemer, M.U.G., Bhatia, S., Cori, A., Nouvellet, P., Herringer, M., et al.,
vaccination. Expert Rev. Vaccines 21 (5), 597–600. 2019. Real-time Epidemic Forecasting: Challenges and Opportunities. Health Secur.
Chavda, V.P., Apostolopoulos, V., 2022 Feb. Global impact of delta plus variant and 17 (4), 268–275.
vaccination. Expert Rev. Vaccines [Internet]. 21 (5), 597–600. https://doi.org/ Elsevier. Elsevier Connect. [cited 2023 Jul 9]. Novel Coronavirus Information Center.
10.1080/14760584.2022.2044800. Available from: Available from: https://www.elsevier.com/connect/coronavirus-information-center.
Chavda, V.P., Apostolopoulos, V., 2022 Oct 3. COVID-19 vaccine design and vaccination El-Shabasy, R.M., Nayel, M.A., Taher, M.M., Abdelmonem, R., Shoueir, K.R., Kenawy, E.
strategy for emerging variants. Expert Rev. Vaccines 21 (10), 1359–1361. R., 2022 Apr. Three waves changes, new variant strains, and vaccination effect
Chavda, V.P., Redwan, E.M., 2022. SARS-CoV-2: Immunopeptidomics and Other against COVID-19 pandemic. Int. J. Biol. Macromol. 15 (204), 161–168.
Immunological Studies. Vaccines. 10 (11). Elwazir, M.Y., Hosny, S., 2020. Artificial Intelligence in COVID-19 Ultrastructure.
Chavda, V.P., Gajjar, N., Shah, N., Dave, D.J., 2021. Darunavir ethanolate: Repurposing J Microsc Ultrastruct. 8 (4), 146–147.
an anti-HIV drug in COVID-19 treatment. Available from Eur. J. Med. Chem. Rep. Ergin, S., Kherad, N., Alagoz, M., 2022 Mar. RNA sequencing and its applications in
[Internet]. 3, 100013. https://www.sciencedirect.com/science/article/pii/S277 cancer and rare diseases. Mol. Biol. Rep. 49 (3), 2325–2333.
2417421000133. Esman, A., Cherkashina, A., Mironov, K., Dubodelov, D., Salamaikina, S., Golubeva, A.,
Chavda, V.P., Vora, L.K., Pandya, A.K., Patravale, V.B., 2021 Nov 1. Intranasal vaccines et al., 2022 Oct 1. SARS-CoV-2 Variants Monitoring Using Real-Time PCR.
for SARS-CoV-2: From challenges to potential in COVID-19 management. Available Diagnostics (Basel). 12 (10), 2388.
from Drug Discovery Today [Internet]. 26 (11), 2619–2636. https://www.sciencedi Faria, N.R., Mellan, T.A., Whittaker, C., Claro, I.M., Candido D da S, Mishra S, et al.
rect.com/science/article/pii/S1359644621003317. Genomics and epidemiology of the P.1 SARS-CoV-2 lineage in Manaus, Brazil.
Chavda, V.P., Kapadia, C., Soni, S., Prajapati, R., Chauhan, S.C., Yallapu, M.M., et al., Science. 2021 May 21;372(6544):815–21.
2022 Apr. A global picture: therapeutic perspectives for COVID-19. Immunotherapy Gerke, S., Minssen, T., Cohen, G., 2020. Ethical and legal challenges of artificial
14 (5), 351–371. intelligence-driven healthcare. Artif. Intell. Healthcare. 295–336.
Chavda, V.P., Prajapati, R., Lathigara, D., Nagar, B., Kukadiya, J., Redwan, E.M., et al., Ghayvat, H., Awais, M., Gope, P., Pandya, S., Majumdar, S., 2021 Jun. ReCognizing
2022 Jun 3. Therapeutic monoclonal antibodies for COVID-19 management: an SUspect and PredictiNg ThE SpRead of Contagion Based on Mobile Phone LoCation
update. Expert Opin. Biol. Therapy [Internet]. 22 (6), 763–780. https://doi.org/ DaTa (COUNTERACT): A system of identifying COVID-19 infectious and hazardous
10.1080/14712598.2022.2078160. Available from: sites, detecting disease outbreaks based on the internet of things, edge computing,
Chavda, V.P., Yao, Q., Vora, L.K., Apostolopoulos, V., Patel, C.A., Bezbaruah, R., et al., and artificial intelligence. Sustain. Cities Soc. 1 (69), 102798.
2022 Oct. Fast-track development of vaccines for SARS-CoV-2: The shots that saved Groneberg, D.A., Hilgenfeld, R., Zabel, P., 2005 Jan 20. Molecular mechanisms of severe
the world. Front. Immunol. 3 (13), 961198. acute respiratory syndrome (SARS). Respir. Res. 6 (1), 8.
Chavda, V.P., Chen, Y., Dave, J., Chen, Z.S., Chauhan, S.C., Yallapu, M.M., et al., 2022 Guhathakurata, S., Kundu, S., Chakraborty, A., Banerjee, J.S., 2021. A novel approach to
Nov 2. COVID-19 and vaccination: myths vs science. Expert Rev. Vaccines [Internet]. predict COVID-19 using support vector machine. Data Science for COVID-19.
21 (11), 1603–1620. https://doi.org/10.1080/14760584.2022.2114900. Available 351–364.
from:
15
Gupta, A.K., Parker, B.M., Priyadarshi, V., Parker, J., 2020. Cardiac Adverse Events With Li, S., Yang, S., Zhou, Y., Disoma, C., Dong, Z., Du, A., et al., 2021. Microbiome Profiling
Remdesivir in COVID-19 Infection. Cureus [Internet]. 2020 Oct 24 [cited 2023 Oct Using Shotgun Metagenomic Sequencing Identified Unique Microorganisms in
18]; Available from: https://www.cureus.com/articles/44072-cardiac-adverse- COVID-19 Patients With Altered Gut Microbiota. Front. Microbiol. 12, 712081.
events-with-remdesivir-in-covid-19-infection. Linardatos, P., Papastefanopoulos, V., Kotsiantis, S., 2021 Jan. Explainable AI: A Review
Hafeez, A., Ahmad, S., Siddqui, S.A., Ahmad, M., Mishra, S., 2020. A Review of COVID- of Machine Learning Interpretability Methods. Entropy 23 (1), 18.
19 (Coronavirus Disease-2019) Diagnosis, Treatments and Prevention. Eur. J. Med. Lionta, E., Spyrou, G., Vassilatis, D.K., Cournia, Z., 2014. Structure-based virtual
Oncol. 4 (2), 116–125. screening for drug discovery: principles, applications and recent advances. Curr.
Haleem, A., Javaid, M., Singh, R.P., Suman, R., 2021. Telemedicine for healthcare: Top. Med. Chem. 14 (16), 1923–1938.
Capabilities, features, barriers, and applications. Available from Sensors Int. Liu, Q., Allamanis, M., Brockschmidt, M., Gaunt, A.L., 2018. arXiv.org. 2018 [cited 2023
[Internet]. 2, 100117. https://www.sciencedirect.com/science/article/pii/ Jul 10]. Constrained Graph Variational Autoencoders for Molecule Design. Available
S2666351121000383. from: https://arxiv.org/abs/1805.09076v2.
Haque, A., Engel, J., Teichmann, S.A., Lönnberg, T., 2017 Aug 18. A practical guide to Liu, Y., Wu, Y., Shen, X., Xie, L., 2021. COVID-19 Multi-Targeted Drug Repurposing
single-cell RNA-sequencing for biomedical research and clinical applications. Using Few-Shot Learning. Frontiers in Bioinformatics [Internet]. 2021 [cited 2023
Genome Med. 9 (1), 75. Jul 9];1. Available from: https://www.frontiersin.org/articles/10.3389/
Hazarika, I., 2020 Apr 17. Artificial intelligence: opportunities and implications for the fbinf.2021.693177.
health workforce. Int. Health 12 (4), 241–245. Luong, A.D., Buzid, A., Vashist, S.K., Luong, J.H.T., 2021 Dec. Perspectives on
He, B., Garmire, L., 2020. Prediction of repurposed drugs for treating lung injury in electrochemical biosensing of COVID-19. Curr. Opin. Electrochem. 30, 100794.
COVID-19. ArXiv. 2020 Mar 30;arXiv:2003.14333v2. Maghdid H, Ghafoor K, Sadiq A, Curran K, Rawat DB, Rabie K. A Novel AI-enabled
Heidari, A., Jafari Navimipour, N., Unal, M., Toumaj, S., 2022 Sep 1. Machine learning Framework to Diagnose Coronavirus COVID-19 using Smartphone Embedded
applications for COVID-19 outbreak management. Neural Comput. Applic. 34 (18), Sensors: Design Study. In 2020. p. 180–7.
15313–15348. Malik, Y.S., Sircar, S., Bhat, S., Ansari, M.I., Pande, T., Kumar, P., et al., 2021 Sep. How
Heidari, M., Mirniaharikandehei, S., Khuzani, A.Z., Danala, G., Qiu, Y., Zheng, B., 2020 artificial intelligence may help the Covid-19 pandemic: Pitfalls and lessons for the
Dec. Improving the performance of CNN to predict the likelihood of COVID-19 using future. Rev. Med. Virol. 31 (5), 1–11.
chest X-ray images with preprocessing algorithms. Int. J. Med. Inf. 1 (144), 104284. Managing COVID-19 in resource-limited settings: critical care considerations - PubMed
Hemdan, E.E.D., Shouman, M.A., Karar, M.E., 2020. COVIDX-Net: A Framework of Deep [Internet]. [cited 2023 Jul 9]. Available from: https://pubmed.ncbi.nlm.nih.gov/
Learning Classifiers to Diagnose COVID-19 in X-Ray Images [Internet]. arXiv; 2020 32321566/.
[cited 2023 Jul 10]. Available from: http://arxiv.org/abs/2003.11055. Markus, A.F., Kors, J.A., Rijnbeek, P.R., 2021 Jan. The role of explainability in creating
Ho, T., Wu, S., Chen, J., Li, C., Hsiang, C., 2007 May. Emodin blocks the SARS trustworthy artificial intelligence for health care: A comprehensive survey of the
coronavirus spike protein and angiotensin-converting enzyme 2 interaction. terminology, design choices, and evaluation strategies. J. Biomed. Inform. 1 (113),
Antiviral Res. 74 (2), 92–101. 103655.
Hojyo, S., Uchida, M., Tanaka, K., Hasebe, R., Tanaka, Y., Murakami, M., et al., 2020 Dec. Mashamba-Thompson, T.P., Crayton, E.D., 2020 Apr. Blockchain and Artificial
How COVID-19 induces cytokine storm with high mortality. Inflamm Regener. 40 Intelligence Technology for Novel Coronavirus Disease 2019 Self-Testing.
(1), 37. Diagnostics. 10 (4), 198.
Huang, Z., Chavda, V.P., Vora, L.K., Gajjar, N., Apostolopoulos, V., Shah, N., et al., 2022. Mbunge, E., 2020. Integrating emerging technologies into COVID-19 contact tracing:
2-Deoxy-D-Glucose and its Derivatives for the COVID-19 Treatment: An Update. Opportunities, challenges and pitfalls. Diabetes Metab. Syndr. 14 (6), 1631–1636.
Front. Pharmacol. 13, 899633. Mbunge, E., Akinnuwesi, B., Fashoto, S.G., Metfula, A.S., Mashwama, P., 2021 Jan.
Huang, M.T., Ghai, G., Ho, C.T., 2004 Oct. Inflammatory Process and Molecular Targets A critical review of emerging technologies for tackling COVID-19 pandemic. Hum
for Antiinflammatory Nutraceuticals. Compr. Rev. Food Sci. Food Saf. 3 (4), Behav Emerg Technol. 3 (1), 25–39.
127–139. McNamara, R.P., Caro-Vegas, C., Landis, J.T., Moorad, R., Pluta, L.J., Eason, A.B., et al.,
Hufsky, F., Lamkiewicz, K., Almeida, A., Aouacheria, A., Arighi, C., Bateman, A., et al., 2020 Nov 3. High-Density Amplicon Sequencing Identifies Community Spread and
2021 Mar 1. Computational strategies to combat COVID-19: useful tools to Ongoing Evolution of SARS-CoV-2 in the Southern United States. Cell Rep. 33 (5),
accelerate SARS-CoV-2 and coronavirus research. Brief. Bioinform. 22 (2), 642–663. 108352.
Hwang, B., Lee, J.H., Bang, D., 2018 Aug. Single-cell RNA sequencing technologies and Mei, X., Lee, H.C., Diao, K yue, Huang M, Lin B, Liu C, et al. Artificial
bioinformatics pipelines. Exp. Mol. Med. 50 (8), 1–14. intelligence–enabled rapid diagnosis of patients with COVID-19. Nat Med. 2020 Aug;
Imran, A., Posokhova, I., Qureshi, H.N., Masood, U., Riaz, M.S., Ali, K., et al., 2020. 26(8):1224–8.
AI4COVID-19: AI enabled preliminary diagnosis for COVID-19 from cough samples Mir, M.M., Mir, G.M., Raina, N.T., Mir, S.M., Mir, S.M., Miskeen, E., et al., 2023 Jul.
via an app. Inform Med Unlocked. 20, 100378. Application of Artificial Intelligence in Medical Education: Current Scenario and
Jiang, L., Wu, Z., Xu, X., Zhan, Y., Jin, X., Wang, L., et al., 2021 Mar 26. Opportunities Future Perspectives. J. Adv. Med. Educ. Prof. 11 (3), 133–140.
and challenges of artificial intelligence in the medical field: current application, Mishra, S., 2020. Bioinformatics Approach for COVID-19 (Coronavirus) Disease
emerging problems, and problem-solving strategies. J. Int. Med. Res. 49 (3). Prevention Treatment and Drug Validation. EJMO [Internet]. 2020 [cited 2023 Jul
Jobin, A., Ienca, M., Vayena, E., Artificial Intelligence: the global landscape of ethics 9]; Available from: https://www.ejmo.org/10.14744/ejmo.2020.97358/.
guidelines. Mittelstadt, B., 2019 Nov 4. Principles alone cannot guarantee ethical AI. Nat Mach
Kedia, P., Anjum, Katarya, R., 2021. CoVNet-19: A Deep Learning model for the detection Intell. 1 (11), 501–507.
and analysis of COVID-19 patients. Applied Soft Computing. 2021 Jun 1;104: Morley, J., Murphy, L., Mishra, A., Joshi, I., Karpathakis, K., 2022 Jan 31. Governing
107184. Data and Artificial Intelligence for Health Care: Developing an International
Khadela, A., Popat, S., Ajabiya, J., Valu, D., Savale, S., Chavda, V.P., 2023. AI, ML and Understanding. JMIR Form Res. 6 (1), e31623.
Other Bioinformatics Tools for Preclinical and Clinical Development of Drug Murdoch, B., 2021 Sep 15. Privacy and artificial intelligence: challenges for protecting
Products. In: Bioinformatics Tools for Pharmaceutical Drug Product Development health information in a new era. BMC Med. Ethics 22 (1), 122.
[Internet]. 2023 [cited 2023 May 22]. p. 255–84. Available from: https://doi.org/ Naik, N., Hameed, B.M.Z., Shetty, D.K., Swain, D., Shah, M., Paul, R., et al., 2022 Mar.
10.1002/9781119865728.ch12. Legal and Ethical Consideration in Artificial Intelligence in Healthcare: Who Takes
Khan, M., Mehran, M.T., Haq, Z.U., Ullah, Z., Naqvi, S.R., Ihsan, M., et al., 2021 Dec. Responsibility? Front Surg. 14 (9), 862322.
Applications of artificial intelligence in COVID-19 pandemic: A comprehensive Nair, P.C., Miners, J.O., 2014 Nov. Molecular dynamics simulations: from structure
review. Expert Syst. Appl. 15 (185), 115695. function relationships to drug discovery. In Silico Pharmacol. 21 (2), 4.
Kim, J.H., Lee, J.H., Koh, E.S., Park, S.W., Jang, A.S., Kim, D., et al., 2013 Apr. Acute Nkwayep, C.H., Bowong, S., Tewa, J.J., Kurths, J., 2020 Nov. Short-term forecasts of the
eosinophilic pneumonia related to a mesalazine suppository. Asia Pac. Allergy 3 (2), COVID-19 pandemic: a study case of Cameroon. Chaos Solitons Fractals 1 (140),
136–139. 110106.
Knowledge Graph-Based Approaches to Drug Repurposing for COVID-19 | Journal of Norgeot, B., Quer, G., Beaulieu-Jones, B.K., Torkamani, A., Dias, R., Gianfrancesco, M.,
Chemical Information and Modeling [Internet]. [cited 2023 Jul 9]. Available from: et al., 2020 Sep. Minimum information about clinical artificial intelligence
https://pubs.acs.org/doi/10.1021/acs.jcim.1c00642. modeling: the MI-CLAIM checklist. Nat. Med. 26 (9), 1320–1324.
Komorowski, M., Celi, L.A., Badawi, O., Gordon, A.C., Faisal, A.A., 2018 Nov. The Norori, N., Hu, Q., Aellen, F.M., Faraci, F.D., Tzovara, A., 2021 Oct 8. Addressing bias in
Artificial Intelligence Clinician learns optimal treatment strategies for sepsis in big data and AI for health care: A call for open science. Patterns (N Y). 2 (10),
intensive care. Nat. Med. 24 (11), 1716–1720. 100347.
Konings, F., Perkins, M.D., Kuhn, J.H., Pallen, M.J., Alm, E.J., Archer, B.N., et al., 2021 Obermeyer, Z., Emanuel, E.J., 2016 Sep 29. Predicting the Future — Big Data, Machine
Jul. SARS-CoV-2 Variants of Interest and Concern naming scheme conducive for Learning, and Clinical Medicine. N. Engl. J. Med. 375 (13), 1216–1219.
global discourse. Nat. Microbiol. 6 (7), 821–823. OECD [Internet]. [cited 2023 Jul 9]. Using artificial intelligence to help combat COVID-
Lalmuanawma, S., Hussain, J., Chhakchhuak, L., 2020 Oct. Applications of machine 19. Available from: https://www.oecd.org/coronavirus/policy-responses/using-
learning and artificial intelligence for Covid-19 (SARS-CoV-2) pandemic: A review. artificial-intelligence-to-help-combat-covid-19-ae4c5c21/.
Chaos Solitons Fractals 1 (139), 110059. Ong, E., Wong, M.U., Huffman, A., He, Y., 2020. COVID-19 Coronavirus Vaccine Design
Lee, U., Kim, A., 2021. Benefits of Mobile Contact Tracing on COVID-19: Tracing Using Reverse Vaccinology and Machine Learning. Frontiers in Immunology
Capacity Perspectives. Frontiers in Public Health [Internet]. 2021 [cited 2023 Jul [Internet]. 2020 [cited 2023 Jul 10];11. Available from: https://www.frontiersin.
10];9. Available from: https://www.frontiersin.org/articles/10.3389/ org/articles/10.3389/fimmu.2020.01581.
fpubh.2021.586615. Pacheco Rocha, N., Dias, A., Santinha, G., Rodrigues, M., Queirós, A., Rodrigues, C., 2019
Lee, C.H., Giuliani, F., 2019 Jul. The Role of Inflammation in Depression and Fatigue. Sep. Smart Cities and Healthcare: A Systematic Review. Technologies. 7 (3), 58.
Front. Immunol. 19 (10), 1696. Park, H., Baek, S., Sen, A., Jung, B., Shim, J., Park, Y.C., et al., 2022 Feb 22.
Ultrasensitive and Selective Field-Effect Transistor-Based Biosensor Created by Rings
of MoS2 Nanopores. ACS Nano 16 (2), 1826–1835.
16
Peiffer-Smadja, N., Maatoug, R., Lescure, F.X., D’Ortenzio, E., Pineau, J., King, J.R., 2020 cities and towns are under quarantine. Infect. Control Hosp. Epidemiol. 41 (7),
May 22. Machine Learning for COVID-19 needs global collaboration and data- 826–830.
sharing. Nat Mach Intell. 2 (6), 293–294. Srinivasan, S., Batra, R., Chan, H., Kamath, G., Cherukara, M.J., Sankaranarayanan, S.K.
Piccialli, F., Di Cola, V.S., Giampaolo, F., Cuomo, S., 2021 Dec. The Role of Artificial R.S., 2021 May 18. Artificial Intelligence-Guided De Novo Molecular Design
Intelligence in Fighting the COVID-19 Pandemic. Inf. Syst. Front. 23 (6), 1467–1497. Targeting COVID-19. ACS Omega 6 (19), 12557–12566.
Pineda, B., Pérez de la Cruz, V., Hernández Pando, R., Sotelo, J., 2021 Jan. Quinacrine as Swayamsiddha, S., Prashant, K., Shaw, D., Mohanty, C., 2021. The prospective of
a potential treatment for COVID-19 virus infection. Eur. Rev. Med. Pharmacol. Sci. Artificial Intelligence in COVID-19 Pandemic. Health Technol (Berl). 11 (6),
25 (1), 556–566. 1311–1320.
Pinzi, L., Rastelli, G., 2019 Sep 4. Molecular Docking: Shifting Paradigms in Drug The Emerging Role of Blockchain Technology Applications in Routine Disease
Discovery. Int. J. Mol. Sci. 20 (18), 4331. Surveillance Systems to Strengthen Global Health Security - ProQuest [Internet].
Pirolli, D., Righino, B., De Rosa, M.C., 2021 Jun. Targeting SARS-CoV-2 Spike Protein/ [cited 2023 Jul 9]. Available from: https://www.proquest.com/openview/
ACE2 Protein-Protein Interactions: a Computational Study. Mol Inform. 40 (6), e38bb6628b98f3f6ad4219ea49741ccc/1?pq-origsite=gscholar&cbl=2061777.
e2060080. Torrente-Rodríguez, R.M., Lukas, H., Tu, J., Min, J., Yang, Y., Xu, C., et al., 2020 Dec 2.
Price, W.N., Cohen, I.G., 2019 Jan. Privacy in the Age of Medical Big Data. Nat. Med. 25 SARS-CoV-2 RapidPlex: A Graphene-Based Multiplexed Telemedicine Platform for
(1), 37–43. Rapid and Low-Cost COVID-19 Diagnosis and Monitoring. Matter. 3 (6), 1981–1998.
Raji, I.D., Buolamwini, J., 2019. Actionable Auditing: Investigating the Impact of Trials.ai – Our Smart Protocol technology helps teams design better clinical trials with AI
Publicly Naming Biased Performance Results of Commercial AI Products. In: [Internet]. [cited 2023 Jul 10]. Available from: https://www.trials.ai/.
Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society Tsopra, R., Fernandez, X., Luchinat, C., Alberghina, L., Lehrach, H., Vanoni, M., et al.,
[Internet]. Honolulu HI USA: ACM; 2019 [cited 2023 Oct 13]. p. 429–35. Available 2021 Oct 2. A framework for validating AI in precision medicine: considerations
from: https://dl.acm.org/doi/10.1145/3306618.3314244. from the European ITFoC consortium. BMC Med. Inf. Decis. Making 21 (1), 274.
Reporting guidelines for clinical trial reports for interventions involving artificial Using Artificial Intelligence to Detect COVID-19 and Community-acquired Pneumonia
intelligence: the CONSORT-AI extension | Nature Medicine [Internet]. [cited 2023 Based on Pulmonary CT: Evaluation of the Diagnostic Accuracy | Radiology
Jul 10]. Available from: https://www.nature.com/articles/s41591-020-1034-x. [Internet]. [cited 2023 Jul 9]. Available from: https://pubs.rsna.org/doi/full/
Reynolds, C.J., Pade, C., Gibbons, .J.M, Otter, A.D., Lin, K.M., Muñoz Sandoval, D., et al. 10.1148/radiol.2020200905.
Immune boosting by B.1.1.529 (Omicron) depends on previous SARS-CoV-2 Vaishya, R., Javaid, M., Khan, I.H., Haleem, A., 2020 Jul 1. Artificial Intelligence (AI)
exposure. Science. 2022 Jul 15;377(6603):eabq1841. applications for COVID-19 pandemic. Diabetes Metab. Syndr. 14 (4), 337–339.
Robson, B., 2020. Computers and viral diseases. Preliminary bioinformatics studies on Vora, L.K., Gholap, A.D., Jetha, K., Thakur, R.R.S., Solanki, H.K., Chavda, V.P., 2023 Jul.
the design of a synthetic vaccine and a preventative peptidomimetic antagonist Artificial Intelligence in Pharmaceutical Technology and Drug Delivery Design.
against the SARS-CoV-2 (2019-nCoV, COVID-19) coronavirus. Comput Biol Med. Pharmaceutics. 15 (7), 1916.
2020 Apr;119:103670. Waman, V.P., Sen, N., Varadi, M., Daina, A., Wodak, S.J., Zoete, V., et al., 2021 Mar 22.
Salas Rojas, M., Silva Garcia, R., Bini, E., Pérez De La Cruz, V., León Contreras, J.C., The impact of structural bioinformatics tools and resources on SARS-CoV-2 research
Hernández Pando, R., et al. Quinacrine, an Antimalarial Drug with Strong Activity and therapeutic strategies. Brief. Bioinform. 22 (2), 742–768.
Inhibiting SARS-CoV-2 Viral Replication In Vitro. Viruses. 2021 Jan 17;13(1):121. Wang X jun, Hu W, Zhang T yu, Mao Y ying, Liu N nan, Wang S qi. Irbesartan, an FDA
Santosh, K.C., 2020 Mar 18. AI-Driven Tools for Coronavirus Outbreak: Need of Active approved drug for hypertension and diabetic nephropathy, is a potent inhibitor for
Learning and Cross-Population Train/Test Models on Multitudinal/Multimodal Data. hepatitis B virus entry by disturbing Na+-dependent taurocholate cotransporting
J. Med. Syst. 44 (5), 93. polypeptide activity. Antiviral Research. 2015 Aug;120:140–6.
Sarmadi, A., Hassanzadeganroudsari, M., Soltani, M., 2023. Artificial Intelligence and Wang, X., Guan, Y., 2021 Jan. COVID-19 drug repurposing: A review of computational
Machine Learning Applications in Vaccine Development. In: Bioinformatics Tools for screening methods, clinical trials, and protein interaction assays. Med. Res. Rev. 41
Pharmaceutical Drug Product Development [Internet]. 2023 [cited 2023 Jul 5]. p. (1), 5–28.
233–53. Available from: https://doi.org/10.1002/9781119865728.ch11. Wang, H., Li, X., Li, T., Zhang, S., Wang, L., Wu, X., et al., 2020 Sep. The genetic
Schlesinger, N., Firestein, B.L., Brunetti, L., 2020 Aug. Colchicine in COVID-19: an Old sequence, origin, and diagnosis of SARS-CoV-2. Eur. J. Clin. Microbiol. Infect. Dis. 39
Drug. New Use. Curr Pharmacol Rep. 6 (4), 137–145. (9), 1629–1635.
Seh, A.H., Zarour, M., Alenezi, M., Sarkar, A.K., Agrawal, A., Kumar, R., et al., 2020 May Wang, L., Zhang, Y., Wang, D., Tong, X., Liu, T., Zhang, S., et al., 2021 [cited 2023 Oct
13. Healthcare Data Breaches: Insights and Implications. Healthcare (Basel). 8 (2), 16];8.. Artificial Intelligence for COVID-19: A Systematic Review. Available from:
133. Frontiers in Medicine [Internet]. https://www.frontiersin.org/articles/10.33
Selvaraj, V., Bavishi, C., Patel, S., Dapaah-Afriyie ,K., 2021. Complete heart block 89/fmed.2021.704256.
associated with Remdesivir in COVID-19: a case report. Abumuaileq RRY, Borovac Wei, S., Weiss, Z.R., Williams, Z., Rapid Multiplex Small DNA Sequencing on the MinION
JA, Ang R, Simovic S, Chakir M, editors. European Heart Journal - Case Reports. Nanopore Sequencing Platform.
2021 Jul 1;5(7):ytab200. Wensel, C.R., Pluznick, J.L., Salzberg, S.L., Sears, C.L., 2022. Next-generation
Shahroz, M., Ahmad, F., Younis, M.S., Ahmad, N., Kamel Boulos, M.N., Vinuesa, R., et al., sequencing: insights to advance clinical investigations of the microbiome. J Clin
2021 Sep. COVID-19 digital contact tracing applications and techniques: A review Invest [Internet]. 2022 Apr 1;132(7). Available from: https://doi.org/10.1172/
post initial deployments. Transport. Eng. 5, 100072. JCI154944.
Shan, B., Broza, Y.Y., Li, W., Wang, Y., Wu, S., Liu, Z., et al., 2020 Sep 22. Multiplexed White, R., Pellefigues, C., Ronchese, F., Lamiable, O., Eccles, D., 2017. Investigation of
Nanomaterial-Based Sensor Array for Detection of COVID-19 in Exhaled Breath. ACS chimeric reads using the MinION [Internet]. F1000Research; 2017 [cited 2023 Jul
Nano 14 (9), 12125–12132. 9]. Available from: https://f1000research.com/articles/6-631.
Shanbehzadeh, M., Kazemi-Arpanahi, H., Nopour, R., 2021 Mar. Performance evaluation Wong, Z.S.Y., Zhou, J., Zhang, Q., 2019 Feb 1. Artificial Intelligence for infectious
of selected decision tree algorithms for COVID-19 diagnosis using routine clinical disease Big Data Analytics. Infection, Disease & Health. 24 (1), 44–48.
data. Med. J. Islam Repub. Iran 1 (35), 29. Woo, P.C.Y., Lau, S.K.P., Chu C ming, Chan, K hung, Tsoi H wah, Huang Y, et al.
Dk S. Sharma DK and Baruah C. Bioinformatics and Artificial Intelligence to Combat Characterization and complete genome sequence of a novel coronavirus, coronavirus
Covid-19 and Beyond. Bioinform Proteom Opn Acc J 2021, 5(1): 000140. 2021. HKU1, from patients with pneumonia. J Virol. 2005 Jan;79(2):884–95.
Sharma, A., Bahl, S., Bagha, A.K., Javaid, M., Shukla, D.K., Haleem, A., 2022. Blockchain Xu, Z., Peng, C., Shi, Y., Zhu, Z., Mu, K., Wang, X., et al. Nelfinavir was predicted to be a
technology and its applications to combat COVID-19 pandemic. Res. Biomed. Eng. potential inhibitor of 2019-nCov main protease by an integrative approach
38 (1), 173–180. combining homology modelling, molecular docking and binding free energy
Shen, X., Tang, H., McDanal, C., Wagh, K., Fischer, W., Theiler, J., et al., 2021 Apr 14. calculation [Internet]. bioRxiv; 2020 [cited 2023 Jul 10]. p. 2020.01.27.921627.
SARS-CoV-2 variant B.1.1.7 is susceptible to neutralizing antibodies elicited by Available from: https://www.biorxiv.org/content/10.1101/2020.01.27.921627v1.
ancestral spike vaccines. Cell Host Microbe 29 (4), 529–539.e3. Yu, R., Chen, L., Lan, R., Shen, R., Li, P., 2020 Aug. Computational screening of
Shirbhate, E., Pandey, J., Patel, V.K., Kamal, M., Jawaid, T., Gorain, B., et al., 2021. antagonists against the SARS-CoV-2 (COVID-19) coronavirus by molecular docking.
Understanding the role of ACE-2 receptor in pathogenesis of COVID-19 disease: a Int. J. Antimicrob. Agents 56 (2), 106012.
potential approach for therapeutic intervention. Pharmacol. Rep. 73 (6), 1539–1550. Yu, M., Tang, A., Brown, K., Bouchakri, R., St-Onge, P., Wu, S., et al., 2021 Dec.
Shorten, C., Khoshgoftaar, T.M., Furht, B., 2021 Jan 11. Deep Learning applications for Integrating artificial intelligence in bedside care for covid-19 and future pandemics.
COVID-19. J. Big Data. 8 (1), 18. BMJ 31 (375), e068197.
Singh, D., Kumar, V., Vaishali null, Kaur M. Classification of COVID-19 patients from Zech, J.R., Badgeley, M.A., Liu, M., Costa, A.B., Titano, J.J., Oermann, E.K., 2018 Nov 6.
chest CT images using multi-objective differential evolution-based convolutional Variable generalization performance of a deep learning model to detect pneumonia
neural networks. Eur J Clin Microbiol Infect Dis. 2020 Jul;39(7):1379–89. in chest radiographs: A cross-sectional study. PLoS Med. 15 (11), e1002683.
Skayem, C., Ayoub, N., 2020 Sep. Carvedilol and COVID-19: A Potential Role in Reducing Zhang, K., Liu, X., Shen, J., Li, Z., Sang, Y., Wu, X., et al., 2020 Jun 11. Clinically
Infectivity and Infection Severity of SARS-CoV-2. Am. J. Med. Sci. 360 (3), 300. Applicable AI System for Accurate Diagnosis, Quantitative Measurements, and
Soares, F., Villavicencio, A., Anzanello, M.J., Fogliatto, F.S., Idiart, M.A.P., Stevenson, Prognosis of COVID-19 Pneumonia Using Computed Tomography. Cell 181 (6),
M., 2020. A novel high specificity COVID-19 screening method based on simple 1423–1433.e11.
blood exams and artificial intelligence [Internet]. medRxiv; 2020 [cited 2023 Jul 9]. Zhang, C., Lu, Y., 2021 Sep. Study on artificial intelligence: The state of the art and future
p. 2020.04.10.20061036. Available from: https://www.medrxiv.org/content/ prospects. J. Ind. Inf. Integr. 1 (23), 100224.
10.1101/2020.04.10.20061036v1. Zhang, H., Saravanan, K.M., Yang, Y., Hossain, M.T., Li, J., Ren, X., et al., 2020. Deep
Soria-Guerra, R.E., Nieto-Gomez, R., Govea-Alonso, D.O., Rosales-Mendoza, S., 2015 Learning Based Drug Screening for Novel Coronavirus 2019-nCov. Interdiscip. Sci.
Feb. An overview of bioinformatics tools for epitope prediction: implications on 12 (3), 368–376.
vaccine development. J. Biomed. Inform. 53, 405–414. Zhang, D.H., Wu, K.L., Zhang, X., Deng, S.Q., Peng, B., 2020 Mar. In silico screening of
Srinivasa Rao, A.S.R., Vazquez, J.A., 2020 Jul. Identification of COVID-19 can be quicker Chinese herbal medicines with the potential to directly inhibit 2019 novel
through artificial intelligence framework using a mobile phone-based survey when coronavirus. J Integr Med. 18 (2), 152–158.
17
Zhang, H tao, Zhang J song, Zhang H hua, Nan Y dong, Zhao Y, Fu E qing, et al. Zhou, Y., Hou, Y., Shen, J., Huang, Y., Martin, W., Cheng, F., 2020 Mar 16. Network-
Automated detection and quantification of COVID-19 pneumonia: CT imaging based drug repurposing for novel coronavirus 2019-nCoV/SARS-CoV-2. Cell Discov.
analysis by a deep learning-based software. Eur J Nucl Med Mol Imaging. 2020;47 6 (1), 1–18.
(11):2525–32. Zhou, Y., Hou, Y., Shen, J., Huang, Y., Martin, W., Cheng, F., 2020 Mar 16. Network-
Zhou, D., Dejnirattisai, W., Supasa, P., Liu, C., Mentzer, A.J., Ginn, H.M., et al., 2021 Apr based drug repurposing for novel coronavirus 2019-nCoV/SARS-CoV-2. Cell Discov.
29. Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine- 6 (1), 14.
induced sera. Cell 184 (9), 2348–2361.e6.
18

1 s2.0 S2590262823000412 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S2590262823000412 Main

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S2590262823000412 Main

Uploaded by

Copyright:

Available Formats

Current Research in Biotechnology 6 (2023) 100159

Contents lists available at ScienceDirect

Current Research in Biotechnology

An amalgamation of bioinformatics and artificial intelligence for COVID-19

Introduction precise COVID data, aiding healthcare professionals in gaining insights

Fig. 2. Sequencing approaches for COVID-19 (created with Biorender.com).

Fig. 3. Role of AI in COVID-19 (created with Biorender.com).

Fig. 6. Instance of AI applications at different stages of COVID-19 (created with Biorender.com).

Limited generalization References

You might also like