TYPE Original Research

PUBLISHED 25 April 2023

DOI 10.3389/fimmu.2023.1027346

Manual cell selection in single

OPEN ACCESS cell transcriptomics using
EDITED BY
Edda Russo,
University of Florence, Italy
scSELpy supports the analysis
REVIEWED BY
Joseph CF Ng,
of immune cell subsets
University College London,
United Kingdom
Carolina Moraes Cabe, Mark Dedden 1, Maximilian Wiendl 1, Tanja M. Müller 1,2, Markus
Institut Pasteur, France
F. Neurath 1,2 and Sebastian Zundler 1,2*
*CORRESPONDENCE
Sebastian Zundler
1
Department of Medicine 1, University Hospital Erlangen and Friedrich-Alexander-Universität
[email protected] Erlangen-Nürnberg, Erlangen, Germany, 2 Deutsches Zentrum Immuntherapie (DZI), University
Hospital Erlangen, Erlangen, Germany
SPECIALTY SECTION
This article was submitted to
Inflammation,
a section of the journal
Frontiers in Immunology Introduction: Single cell RNA sequencing plays an increasing and indispensable
RECEIVED 24 August 2022 role in immunological research such as in the field of inflammatory bowel
ACCEPTED 07 April 2023 diseases (IBD). Professional pipelines are complex, but tools for the manual
PUBLISHED 25 April 2023
selection and further downstream analysis of single cell populations are missing
CITATION so far.
Dedden M, Wiendl M, Müller TM,
Neurath MF and Zundler S (2023) Manual
cell selection in single cell transcriptomics Methods: We developed a tool called scSELpy, which can easily be integrated
using scSELpy supports the analysis of into Scanpy-based pipelines, allowing the manual selection of cells on single cell
immune cell subsets.
transcriptomic datasets by drawing polygons on various data representations.
Front. Immunol. 14:1027346.
doi: 10.3389/fimmu.2023.1027346 The tool further supports the downstream analysis of the selected cells and the
COPYRIGHT
plotting of results.
© 2023 Dedden, Wiendl, Müller, Neurath and
Zundler. This is an open-access article Results: Taking advantage of two previously published single cell RNA
distributed under the terms of the Creative
Commons Attribution License (CC BY). The sequencing datasets we show that this tool is useful for the positive and
use, distribution or reproduction in other negative selection of T cell subsets implicated in IBD beyond standard
forums is permitted, provided the original
clustering. We further demonstrate the feasibility for subphenotyping T cell
author(s) and the copyright owner(s) are
credited and that the original publication in subsets and use scSELpy to corroborate earlier conclusions drawn from the
this journal is cited, in accordance with dataset. Moreover, we also show its usefulness in the context of T cell receptor
accepted academic practice. No use,
distribution or reproduction is permitted
sequencing.
which does not comply with these terms.
Discussion: Collectively, scSELpy is a promising additive tool fulfilling a so far
unmet need in the field of single cell transcriptomic analysis that might support
future immunological research.

KEYWORDS

single cell RNA sequencing, transcriptomics, chronic inflammation, inflammatory bowel

disease, gut homing

Frontiers in Immunology 01 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

Introduction sequencing data might be a helpful additive tool for exploring the
dataset. Although there are already some interactive single cell
Progress in life sciences has led to deep insights into analysis tools, which are very user-friendly and allow to manually
physiological and pathological processes in recent years. However, select cells by polygons without the need for much bioinformatics
in turn, this also resulted in more and more sophisticated problems knowledge (19, 20), these applications are limited, when it comes to
to be further addressed. In particular, this concerns the integrating other tools, changing analysis parameters or performing
inconceivable complexity of immunological processes that sophisticated downstream analyses. On the other hand, despite a
research has discovered and now tries to understand in even huge variety of solutions for these advanced issues in Scanpy, there
greater detail. is currently no easy way to integrate manual cell selection by
One example are immune-mediated inflammatory disorders polygons in Scanpy-based pipelines. Collectively, there is an
(IMIDs) (1) such as the inflammatory bowel diseases (IBD) unmet need for a tool facilitating not only manual cell selection
ulcerative colitis (UC) and Crohn’s disease (CD) characterized by on single cell data, but also supporting downstream characterization
relapsing-remitting chronic inflammation of the gastrointestinal tract and analyses of the selected population, e.g. in terms of (differential)
(2, 3). While an important role of the immune system in the gene expression, in Scanpy.
pathogenesis of IBD had early been demonstrated (4), further Thus, we aimed to close this gap and to simplify the analysis of
ground-breaking insights in fields such as genetics (5) have shaped immune cell subsets in scRNA-seq analyses in the context of
the current picture of multifactorial diseases driven by antigen chronic inflammation. To offer Scanpy users the ability
translocation from a dysbiotic luminal microenvironment over a to annotate their cells of interest by means of manual
leaky epithelial barrier into the lamina propria of a selection, we developed scSELpy (single cell SELection python).
genetically susceptible individual (6). Here, dysregulated immune In addition to selecting cells of interest by drawing polygons
responses are evoked and cause inflammation resulting in tissue around them, it supports further downstream analysis and the
destruction and a vicious cycle of further host-environment generation of publication-ready plots. Our data show that
miscommunication (7). This also triggers the recruitment of our tool is useful to analyze immune cell heterogeneity in IBD
immune cells such as T lymphocytes from the peripheral blood in scRNA-seq beyond conventional clustering and might
to the intestinal tissue, a process called gut homing that is specifically therefore become an important application for future single cell
regulated in the gut and involves molecules such as the chemokine transcriptomic analyses.
receptor CX3CR1 or the integrin a4b7 (8). This also contributes to
the expansion of diverse pro-inflammatory effector and effector
memory T cell subsets in the inflamed gut including T helper 1 Materials and methods
(TH1) and T helper 17 (TH17) cells (9), while regulatory T cells
cannot sufficiently suppress these disease-driving cells (10). scSELpy
Technological advances and also the introduction of novel
therapeutic approaches (11–13) into the clinics have led to scSELpy was developed as a Scanpy extension and solely uses
unprecedented insights into the regulation of aberrant immune libraries required by Scanpy such as matlibplot (21) and numpy
responses in chronic inflammation in general and IBD in particular. (22) (Table 1). It allows the scSELpy user to select cells by drawing
In this context, the single cell RNA-sequencing (scRNA-seq) polygons on top of scatter plots or on either of the following
technology has become a popular and indispensable technique for dimension-reduced representations: UMAP (23), TSNE (24), PCA
interrogating the transcriptome on single cell level and for resolving or other Scanpy-supported embeddings, limited to 2 dimensions.
the heterogeneity of various subsets of adaptive and innate immune The selected cells will be annotated according to the names given to
cells involved in inflammatory networks in chronic inflammation the drawn polygons that they are located within, separated by
(14). Consistently, while the first single cell being sequenced was comma. Subsequently, this cell annotation will be stored as
reported in 2009 (15), more than 4000 Pubmed-indexed articles observations in the Anndata object (25). The coordinates of the
published in 2021 mention single-cell sequencing. polygon itself will be stored as unstructured data. Polygons are
To aid in the interpretation of these big data, many tools have denominated as integers by default, converted to string. A scSELpy
been written over the past years to support scRNA-seq analyses function allows the user to easily convert the default names saved in
(16). Two programs are the backbone for most of these analyses: the the annotations to custom names.
Python library Scanpy (17) and the R package Seurat (18). To detect The invocation function of scSELpy mimics Scanpy’s plotting
communities of cells with similar features, unsupervised clustering function to make it easy for the user to switch between using Scanpy
according to the Leiden or Louvain algorithms can easily be and scSELpy. The scSELpy tool accepts all Scanpy parameters,
integrated into Scanpy- or Seurat-based pipelines. However, these except for the “Layer” parameter and offers additional parameters
approaches may sometimes be limited or time-consuming, when for fine tuning and re-plotting of polygons.
the goal is to analyze a specific subset of cells that does not directly Upon invocation of scSELpy, it will determine if the user is
match to the clusters identified. running Python in a shell or in a notebook environment. For
We hypothesized that in these situations, manual selection of scSELpy to work on a notebook, it switches to an interactive
cells on any two-dimensional representation of single cell matplotlib plotting backend. After the cell selection has been

Frontiers in Immunology 02 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

TABLE 1 Version list.

Name Language Version scSELpy import

scselpy Python 1.0.0 –

scanpy Python 1.7.2 Yes

numpy Python 1.21.1 Yes

matplotlib Python 0.11.6 Yes

pandas Python 1.2.4 No

scipy Python 1.6.2 No

jupyter Python 6.4.12 No

rpy2 Python 3.4.5 No

scirpy Python 0.11.2 No

magic-impute Python 3.0.0 No

sklearn Python 1.0.2 No

scran R 1.14.6 No

conducted, it will switch back to the default matplotlib plotting Data analysis
backend. On a Python shell such as ipython, a backend switch will
not be conducted. We analyzed two scRNA-seq datasets in this manuscript, which
Afterwards, scSELpy will call Scanpy to create a plot of the have been previously published under GSE162624 (26) and
specified embedding. While the plot is open, matplotlib’s function GSM6346300. The data were preprocessed and normalized in the
“ginput” is called, which will catch the coordinates of all same way as in the original study of GSE162624. The entire analysis
mouseclicks on the plot. When the user is finished selecting the was conducted on Jupyter notebook v6.4.12. Cells were selected and
coordinates of a single polygon, the coordinates are sent to the plot annotated using scSELpy. All plots were generated using Matplotlib,
function of matplotlib.pyplot, in order to draw the polygon on the Scanpy and scSELpy. All data imputations were conducted by
Scanpy generated plot. After all polygons are drawn, scSELpy will MAGIC (27). T cell receptor analyses were performed with
switch back to a static backend if necessary. Subsequently it will call Scirpy (28).
Scanpy and Matplotlib again to generate a final image for output.
For each polygon the contains_points function of Matplotlib is
called, in order to determine which cells are located within which Gene enrichment calculation
polygon. This function tests if a given cell coordinate of the passed
embedding is located within the given polygon. The output are x Assuming that one Unique Molecular Identifier (UMI)
boolean lists, where x is the amount of drawn polygons. These represents one detected mRNA transcript, the transcripts per
boolean lists are converted by scSELpy to a single list that contains, million (TPM) for a given gene were calculated by dividing all
which cell is located in which polygon. The list is assigned to an UMIs of this gene in a specific population by all UMIs in the same
observation in the anndata object, which is updated in place. population, multiplied by one million.
Additionally, scSELpy has a three functions for calculating the i)
percentage of cells in each cluster or region, ii) percentage of cells UMI count for given gene
TPM = ( )*106
expressing a given gene in each cluster or region, iii) transcripts per Total UMI count
million (TPM) of a given gene in each cluster or region (Table 2). The enrichment for a given gene was calculated by dividing the
These functions can be used on any observation of the Anndata TPM of a gene within a specific population by the TPM withing all
object and is therefore not exclusive to scSELpy generated regions. cells outside that population.

TABLE 2 Supported read-outs for manual cell selections with scSELpy – example from Figure 2B.

Percentage of cells Percentage of cells expressing TPM of

CCR7 SELL CCR7 SELL

Selection 62.57 61.21 57.77 267.51 284.72

Other cells 37.43 19.38 25.19 61.81 107.08

Frontiers in Immunology 03 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

TPM of all cells in specific population highest enrichment, we obtained only a 50.74-fold increase in
Enrichment =
TPM of all other cells outside of the population CX3CR1 TPM. Collectively, these data suggested that in specific
scenarios, polygon gates drawn with scSELpy result in more specific
positive selection of cell populations enriched for a certain gene
than conventional clustering.
Gaussian mixture model

Cells were divided into groups using a gaussian mixture model

with k-means as initializer based on the normalized mRNA-derived
Negative selection with scSELpy
UMI count of two marker genes. This was done in Python with
In a next step, we aimed to show that our application is also
sklearn, using the GaussianMixture function from sklearn.mixture
useful for the negative selection of cells of interest. Specifically, we
and the Kmeans function from sklearn.cluster.
sought to analyze T cell subsets relevant in the inflamed mucosa in
IBD. Thus, since the dataset comprised peripheral blood memory T
cells expressing the gut-homing marker a4b7, we aimed to identify
Availabilty of scSELpy
T cells homing to the lamina propria by excluding central memory
T (TCM) cells, which home to the gut-associated lymphoid tissue
scSELpy is available for download at https://github.com/
(29). To this end, we used polygon gates to mark regions high in
MarkDedden/scSELpy together with installation instructions, the
expression of the TCM markers CCR7 and SELL (CD62L) on
data analysis pipeline and a link to the documentation, which
UMAP plots (Figure 2A). The overlay of regions, where both
includes a tutorial.
genes were expressed in high levels was removed to obtain the
cells equipped for access to the inflamed lamina propria (Figure 2B)
as evident by a low prevalence of cells expressing CCR7 and/or
Results SELL and low expression of these genes in the retained region
(Table 2). The remaining cells were further re-analyzed from raw
Single cell selection in Python
data, creating a new UMAP plot (Figure 2C).
To confirm that the selected region was also enriched for cells
To overcome the problem that standard scRNA-seq analysis
co-expressing CCR7 and SELL, we employed scSELpy on scatter
pipelines, for example based on Scanpy, do not include tools to
plots to “gate” for cells highly expressing CCR7 and SELL
support manual cell selection from scatter or dimension reduction
(CCR7+SELL+), highly expressing CCR7 or SELL (CCR7+SELL-,
plots for further downstream analysis, we developed scSELpy as
CCR7-SELL+) and expressing CCR7 and SELL at low levels or not
detailed in the Methods section.
at all (CCR7-SELL-; Figure 2D). Subsequently, we depicted the
presence of the cells from these four categories in density plots
(Figure 2E). Here, the vast majority of CCR7+SELL+ cells were
Positive selection with scSELpy
located in the removed region, while most of the CCR7-SELL- cells
plotted to the region that was kept, indicating that our tool had
To explore and to demonstrate the functionality of scSELpy, we
helped to correctly eliminate TCM cells.
reanalyzed data (GSE162624) from a previous study (26), where
As scRNA-seq suffers from drop-out events, where mRNA-
CD3+CD4+CD45RO+a4+b7+ gut-homing memory T cells from the
transcripts that are present in a cell might not be detected, we
peripheral human blood were purified by fluorescence-activated cell
employed data imputation using MAGIC (27) to verify that the
sorting (FACS) and submitted to single cell transcriptomics.
selections made with scSELpy are not excluding cells falsely
Based on a UMAP expression plot for CX3CR1 generated by
negative for CCR7 and/or SELL. We overlayed the polygons of
Scanpy, we selected a region high in cells expressing CX3CR1
Figure 2A on the imputed data (Figure S1A) and further depicted
(Figures 1A–D). Indeed, a more than 90-fold increase in CX3CR1
the imputed data in a scatter plot and density plots (Figures S1B, C).
TPM were detected in the selected cells compared to the non-
Indeed, the enrichment of CCR7+SELL+ in the overlap of the
selected cells (Figure 1E, Table 3). Importantly, the region high in
polygons was maintained, supporting the notion that manual
CX3CR1 selected by the polygon gate was different from the clusters
selections predominantly include false negative cells (which is
generated by the Leiden community detection algorithm with a
intended), but do not exclude them to a relevant degree.
resolution of 0.5 (Figure 1F) and the enrichment of CX3CR1
transcripts in the manual selection compared with non-selected
cells was higher than when comparing Leiden clusters with high vs.
low CX3CR1 expression (Figure 1G), where the enrichment was
scSELpy allows for subphenotyping
only below 40-fold increase in CX3CR1 TPM. To investigate if
of T cells
further unsupervised subclustering would lead to an enrichment To confirm that the cells remaining after negative selection
comparable to manual selection, we subclustered cluster 6 (Figure 2C) included tissue-homing TEM cells relevant in IBD and
(Figure 1H) and calculated the enrichment for each subcluster. to characterize them, we explored the expression of TBX21 and
Even when combining the three subclusters (6,0, 6,2, 6,4) with the RORC as key transcription factors for TH1 and TH17 cells,

Frontiers in Immunology 04 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

A B

C D

E F

G H

FIGURE 1
Positive selection of cell subsets on a Scanpy-generated UMAP plot with the scSELpy tool. (A) UMAP expression plot of CX3CR1. Regions with high
CX3CR1 expression were manually selected using scSELpy. (B–D) The cells located in the selected region can be highlighted (B), removed (C), or
isolated (D) in downstream analysis. (E) Barplot with CX3CR1 transcripts per million transcripts (TPM) for the selected cells versus all other cells.
(F) Leiden clustering of the dataset with a resolution of 0.5. (G) Barplot with CX3CR1 TPM for cluster 6 and all the other clusters together.
(H) Subclustering of the leiden cluster 6 from (E).

respectively (30, 31). TBX21 and RORC were expressed in substantially increased in the selected cell clusters (Figure 3D).
overlapping regions that were manually selected by polygons This was consistent with the notion that TBX21+RORC-, TBX21-
(Figure 3A) resulting in the four populations of TBX21-RORC-, RORC+ and TBX21+RORC+ cells corresponded to TH1, TH17 and
TBX21 + RORC - , TBX21 - RORC + and TBX21 + RORC + cells TH1/17 cells (a subset that has been described in the gut of patients
(Figure 3B). Interestingly, this matched well to Leiden clustering with CD (32)), respectively. Thus, we further aimed to corroborate
at a resolution of 0.5 (Figure 3C). Again, data imputation with successful T helper cell subset identification by our tool and
MAGIC retrospectively supported the chosen manual selection analyzed chemokine receptor expression in these three datasets.
(Figure S1D). As expected based on previous reports (33), TBX21+RORC- cells
To validate enriched expression of TBX21 and RORC in the were CXCR3high, but CCR4low and CCR6low, TBX21-RORC+ cells
selected populations, we calculated the TPM of the two genes. CCR6high and CCR4high, but CXCR3low and TBX21+RORC+ cells
Indeed, mRNA expression of both transcription factors was were CXCR3high and CCR6high, but CCR4low (Figure 3E).

Frontiers in Immunology 05 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

TABLE 3 Enrichment of CX3CR1. resistant Tregs were clearly enriched in the Treg fraction that did not
bind vedolizumab (Figure 4D, E). Taken together, these data showed
Cluster Enrichment [fold] that scSELpy is able to reproduce earlier findings and is, thus, a valid
6 38.3 tool for advanced analyses of single cell transcriptomics in general
6,0 45.3
and in the context of IBD in particular.

6,1 1.2

6,2 13.7 scSELpy helps analyzing protein expression

and TCR sequencing data
6,3 4.3

6,4 16.6 To demonstrate that scSELpy is able to analyze protein

6,0 + 6,2 + 6,4 50.7 expression identified by antibody-oligo capture, we analyzed the
dataset GSM6346300 with scSELpy (36). In that dataset, PBMCs
scSELpy selection 91.1
from four patients were isolated from the blood and sequenced on a
10x Chromium controller. This included V(D)J single-cell T cell
Moreover, to further compare the capability of scSELpy to select receptor (TCR) sequencing and Feature Barcoding to capture the
cells enriched for a certain gene with another technique that is protein expression of CD4, CD8 and CD45RA. We loaded the TCR
independent of spatially represented clusters or regions, we applied data using Scirpy, merged it with the scRNA-seq dataset and
a Gaussian mixture model clustering with K-means as initializer for removed all cells that had no TCR detected using the “has_ir”
TBX21 and RORC on the cells obtained in Figure 2C. However, in observation created by Scirpy. We proceeded under the assumption
this case, the algorithm was not able to identify a specific regions that the remaining subset consists of only T cells. We used Leiden
enriched for TBX21 and/or RORC (Figure S2A), strongly clustering with a resolution of 0.5 to assign clusters on a two
contradicting the impression supported by Leiden clustering and dimensional UMAP plot (Figure 5A).
scSELpy selection that T helper cell populations cluster in defined Using scSELpy, we drew a gate for CD8+ cells and a gate for
regions. Thus, this scenario identified another situation, where CD4+ cells on a scatter plot of raw CD8 and CD4 antibody
scSELpy was better suited than established alternative solutions. interaction-derived UMI counts with a logaritmic axis
Collectively, these data showed that our tool had indeed (Figure 5B). We highlighted the CD3D (Figure 5B), CD8A, CD8B
identified TEM cells and was also able to further sub-cluster well- and CD4 mRNA-dervied UMI count (Figure S2B) to show that the
defined T helper cell subsets, which are relevant in IBD. mRNA expression of CD8 and CD4 is enriched in their respective
gates. The CD4+CD8-, CD4-CD8+, CD4+CD8+ and CD4-CD8-
populations were now mapped back to the initial UMAP plot
scSELpy analysis confirms the phenotype (Figure 5C, Figure S2C). Interestingly, in this case, these
of vedolizumab-resistant regulatory T cells populations were well aligned with those identified by the
gaussian mixture model clustering with K-means as initializer
In a next step, we wanted to find out, what kind of cells were (Figure S2D). To demonstrate that this is a feasible starting point
represented in the TBX21-RORC- cluster. Since, regulatory T (Treg) for further downstream analyses and that scSELpy can also assist in
cells are another important gut-homing T cell population and play an understanding TCR sequencing data, we manually selected a region
important role in the resolution of intestinal inflammation (8, 10), we of cluster 5 enriched for CD8+ T cells with scSELpy (Figure 5C).
explored the expression of the key Treg transcription factor FOXP3. Building on the Scanpy-based library for T cell receptor-sequencing
Not very surprising, we detected markedly enriched FOXP3 data analysis, Scirpy, we now plotted all other cells sharing a
expression and manually selected the subset for downstream clonotype with any of the cells in this area and found that these
analysis (Figure 4A). In further support of the Treg nature of these cells mainly map to the clusters 1, 3, 4 and 8 (Figure 5D).
cells, several other Treg-associated molecules were expressed to To further demonstrate the potential to use scSELpy for
significantly higher levels than in the other T cell subsets (Figure 4B). analyzing combined TCR and scRNA sequencing data, we used
Since the dataset consisted of sorted cells binding or not binding Scirpy’s repertoire_overlap function to plot the T cell repertoire
the anti-a4b7 integrin antibody vedolizumab, which is used for overlap between the CD4+CD8- gate and the CD4+CD8+ cells.
clinical therapy of IBD (34, 35), and we had previously shown in Using scSELpy on this plot, we selected clonotypes that are
that dataset that a subset of Tregs is enriched for vedolizumab- present in cells of both gates (Figure 5E) and plotted them in the
resistant cells (26), we now explored whether scSELpy-based analysis CD8 versus CD4 scatterplot (Figure 5F), keeping the gates as set in
comes to the same conclusion. Thus, we depicted vedolizumab Figure 5B, and in the UMAP plot (Figure 5G) as potential starting
binding cells and vedolizumab non-binding cells in our Treg points for further selection and analysis procedures.
population. Consistent with our previous analysis, a majority of Collectively, these approaches demonstrated that scSELpy can
Tregs did not bind vedolizumab (Figure 4C). Moreover, ITGB1 and be used as a convenient and flexible tool helping in the exploration
PI16, two genes that had been found to mark those vedolizumab- and analysis of multi-dimensional single cell sequencing data.

Frontiers in Immunology 06 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

B C

D E

FIGURE 2
Selection and removal of central memory T cells from the dataset using scSELpy to obtain effector memory T cells. (A) Selection of regions enriched
for cells expressing CCR7 and SELL (CD62L, L-Selectin) on UMAP plots. (B) Identification and removal of the cells located in both the CCR7 and SELL
polygon as selected in A (green). (C) Re-analysis of the remaining cells from raw data. Normalization and UMAP dimension reduction is performed
anew. (D) Scatter plot of all cells from the dataset with normalized CCR7 and SELL expression on the x-axis and y-axis, respectively. On this scatter
plot, scSELpy was employed to categorize cells with or without expression of CCR7 and/or SELL. (E) UMAP density plots highlighting the cells
belonging to the categories created in (D). The parameter “vmin” to control the lower limit in the color scale was set to 0.05.

Discussion used in multicolor flow cytometry as the most widely used

technique to interrogate protein expression on single cell level
Single cell RNA sequencing has revolutionized immunological (37). In flow cytometry, such gating serves to select certain cell
research and has helped to substantially increase the resolution of populations in a hierarchical manner and to determine the
immune cell phenotyping (14). However, this comes at the cost of abundance and phenotype of cell subsets in this way (38).
complex in silico analyses to be performed. Manual cell selection in scatter plots visualizing the expression
Here, we introduce a new tool called scSELpy designed to enable of two different genes per cell as offered by scSELpy comes most
the manual selection of cells analyzed by single cell transcriptomics closely to this function. However, the high number of parameters
in scatter plots or dimension reduction representations to allow analyzed by single cell RNA sequencing also imposes the necessity
further downstream analysis. As such, it is inspired by the “gating” to include dimension reduction techniques to appropriately

Frontiers in Immunology 07 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

B C

FIGURE 3
Identification and analysis of effector memory T cell subsets with scSELpy. (A) Selection of regions enriched in cells expressing the T cell
transcription factors TBX21 or RORC among effector memory T cells in the cells obtained in Figure 2C. (B) UMAP plot with cells annotated by
scSELpy based on the selections made in (A). (C) Clustering of the cells based on the leiden algorithm with a resolution of 0.5. (D) Barplot showing
the TBX21 and RORC transcripts per million transcripts (TPM) for the populations identified in (B). (E) CXCR3, CCR6 and CCR4 TPM in the indicated
regions.

visualize and analyze relationships between the single cells (23, 24). expression below the detection threshold of single cell RNA
Consistently, scSELpy also offers manual cell selection on sequencing (39). Our findings with imputed data further support
dimension-reduced UMAP or t-SNE plots. Yet, it needs to be this notion, since the polygons drawn before data imputation
considered that in this case “gating” will not result in the binary captured the majority of false-negative cells and missed only few
selection of cells with and without expression of one or more genes of them. Taken together, these aspects emphasize that, while similar
(or high or low expression), but only in the selection of a population in handling, “gating” on scRNA-seq data, is clearly different from
enriched in cells expressing those genes. This population will also flow cytometry gating.
include closely related cells, in which expression of the gene has not Importantly, scSELpy also supports the analysis of single cell
been detected, which might be due to absent expression or data including TCR sequencing or sequencing of surface proteins

Frontiers in Immunology 08 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

A B C

FIGURE 4
Selection and analysis of regulatory T cells with scSELpy. (A) Selection of the region enriched for FOXP3 expression, using the UMAP representation
presented in Figure 2C. The selections for RORC and TBX21 from Figure 3A are marked with dotted lines. (B) Violin plots of the expression of eight
genes associated with Tregs in the four regions determined based on TBX21 and RORC expression. (C) Depiction of cells binding vedolizumab or
not as determined by FACS sorting prior to sequencing on the population selected in (A). (D) Expression of ITGB1 and PI16 in the cell population
selected in (A). (E) Stacked violin plot of PI16 and ITGB1 expression for Vedolizumab-binding cells and cells that do not bind Vedolizumab.

detected by antibody-oligo reaction. Thus, in the specific case of Several publicly available applications such as the Loup Browser
protein expression analysis, scSELpy can actually be used for binary offered by 10x Genomics, CELLxGENE, the UCSC Cell Browser
gating of surface expression markers very similar to flow cytometry. (19, 20) or Shiny-based applications for scRNA-seq such as
With regard to TCR analyses, we show that scSELpy can identify SCHNAPPs (40) are graphical user interface platforms for single
and plot clonotypes in various representations and might thus help cell analysis, of which the first three mentioned offer similar tools
to explore their role and function. Again, this is not an exclusive for manual cell selection. However, while those applications are easy
feature of scSELpy, since for instance clonotype overlap in different to handle also for researchers without training in bioinformatics, a
cell types can also be identified using Scirpy’s clonotype_imbalance limitation of them is that further downstream analyses are not
function. Thus, it is important to understand scSELpy as a supported. Thus, scSELpy has been designed for use on the Scanpy
complementary tool that can be used together with other platform, one of the standard applications used for the analysis of
important approaches to reach a deeper understanding of the data. single cell RNA-seq data (17) and integrates a workflow to enable

Frontiers in Immunology 09 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

A B

C D

E F

FIGURE 5
Applying scSELpy on T cell receptor (TCR) sequencing and antibody capture data. The single cell RNA sequencing data of PBMCs in blood from
patients treated with Immune checkpoint inhibitors was retrieved from GSM6346300. Only cells, for which a TCR is detected are used in these plots.
(A) UMAP plot of the cells; clusters are defined by leiden with a resolution of 0.5. (B) Scatter plot with the detected CD4 and CD8 antibody UMIs on
the x-axis and y-axis, respectively. Gates set with scSELpy split the data up into CD4+CD8+ double positive cells, single positive CD4+CD8- cells,
single positive CD4-CD8+ cells and CD4-CD8- double negative cells. The CD3D mRNA derived UMI count is highlighted in red. (C) Visualization of
the four groups defined in (A) on the UMAP plot. A group of cells is selected with scSELpy. (D) UMAP representation of T cells selected in (C)
(magenta) together with all other T cells also expressing one of the TCRs present in the selection (blue). Cells colored in grey do not have a
clonotype that matches a T cell in the selected region. (E) TCR repertoire of CD4+CD8- cells plotted against TCR repertoire of CD4+CD8+ cells for
each clone ID using Scirpy. Clone IDs that are present in both cell populations were selected using scSELpy. (F) Mapping of the cells selected in
(E) on the scatter plot of CD4 and CD8 antibody UMIs with the gates kept as in (B). (G) UMAP representation of the T cells selected in (E).

easy downstream phenotyping of selected cells such as further sub- algorithms dissect the overall cell population into several clusters,
clustering, re-plotting or differential expression analysis. which can subsequently be extracted and further analyzed. It is
A standard technique to identify specific cell populations on single important to mention that scSELpy is not at all meant to replace
cell transcriptomic data is clustering according to the Leiden or such clustering-based identification and selection of cell populations,
Louvain algorithm (41). Depending on the resolution used, these but as an additional tool that might be helpful in certain situations.

Frontiers in Immunology 10 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

One key difference to scSELpy is that Leiden or Louvain might broadly support and facilitate single cell transcriptomic
clustering are unsupervised and thus unbiased. In consequence, analyses for many researchers in the field of immunology in
one might mention that scSELpy unnecessarily introduces bias into general and in IBD in special.
single cell RNA sequencing analysis by allowing to select cells based
on one or more deliberately chosen genes. While this has to been
accepted as a potential limitation and to be kept in mind during Data availability statement
analysis, it is also important to note that conventional clustering
might not always perfectly capture biological processes that are not The datasets presented in this study can be found in online
dominating the phenotype of cell subsets or are shared between repositories. The names of the repository/repositories and accession
subsets. For example, this might be processes of cell migration such number(s) can be found in the article/Supplementary Material.
as gut homing. Consistently, our data show that manual cell
selection by scSELpy helps to increase the enrichment of specific
gene expression in the selected populations. This might be of Author contributions
particular value in a time, where the re-analysis of previously
published datasets from a novel perspective is becoming more MD developed scSELpy and performed the analyses. MD and
and more important (42). SZ designed the analyses and interpreted the data together with
It is essential to underscore that in many situations (e.g. as MW, TM and MN. MD and SZ drafted the manuscript; All authors
documented in Figure 3C or Figure S2D) there exist conventional contributed to the article and approved the submitted version.
alternatives such as Leiden or Louvain clustering or k means
clustering that lead to similar results as manual cell selection with
scSELpy. Thus, in these situations, the benefits (fast, easy) and the Funding
limitations (subjectivity) associated with the use of scSELpy must be
carefully weighed. However, as we show (Figure 1E-H, Figure S2A), This work was supported by the German Research Foundation
there are also scenarios, where scSELpy results in superior selection (DFG, ZU377/4-1), the Else Kröner-Fresenius-Stiftung
of enriched populations. Similarly, one can also consider situations, (2021_EKCS.23) and a DAAD scholarship to MD.
where the use of scSELpy is limited by very rare expression of a gene
or equal distribution over the dimensionality-reduced space and
alternative ways of cell selection will be more helpful. In Acknowledgments
consequence, we think that scSELpy is a valuable part of the
toolbox in state-of-the-art single cell analysis that should The research of TM, MN and SZ was supported by the
especially be used in situations, where conventional community Interdisciplinary Center for Clinical Research (IZKF) and the ELAN
detection or cell type identification are not possible, sub-optimal or program of the Universität Erlangen-Nürnberg, the Fritz-Bender-
very time-consuming or where the role of a gene regardless of the Stiftung, the Ernst Jung-Stiftung, the Else Kröner-Fresenius-Stiftung,
association to a specific (sub-)community is explored. the Thyssen-Stiftung, the German Crohn’s and Colitis Foundation
We demonstrate the feasibility of our approach in a dataset (DCCV), the DFG topic program on Microbiota, the Emerging Field
characterizing gut-homing memory T cells from the peripheral Initiative, the DFG Collaborative Research Centers 643, 796, 1181 and
blood, a cell population that is of particular interest for the TRR241, the Rainin Foundation and the Litwin IBD Pioneers program
pathogenesis of IBD and has become a therapeutic target by of the Crohn’s and Colitis Foundation of America (CCFA). We
blocking its gut homing through the anti-a4b7 integrin antibody acknowledge financial support by the German Research Foundation
vedolizumab (7). In addition to proving the suitability of scSELpy (ZU377/4-1) and Friedrich-Alexander-Universität Erlangen-Nürnberg
for appropriate positive and negative selection strategies, we also within the funding programme “Open Access Publication Funding”.
show that our tool is helpful in supporting the analysis of cell
populations such as TH1, TH17 or TH1/17 cells, all of which have
been demonstrated to crucially implicated in IBD (30–32). Thus, Conflict of interest
scSELpy might help to obtain further insights into immune cell
regulation in IBD and other chronic inflammatory diseases in the The authors declare that the research was conducted in the
future. Moreover, in regulatory T cells our tool was able reproduce absence of any commercial or financial relationships that could be
the earlier finding that a4b7-expressing regulatory T cells are construed as a potential conflict of interest.
enriched in cells “resistant” to vedolizumab and that b1 integrin
and PI16 are highly expressed in those cells (26). It might therefore
also be employed for future translational studies in the field of IBD Publisher’s note
aiming at dissecting the mechanisms of state-of-the art treatment
options at higher resolution. All claims expressed in this article are solely those of the authors
Taken together, to the best of our knowledge, scSELpy is the and do not necessarily represent those of their affiliated
first tool that can offer Scanpy-based manual cell selection. Based on organizations, or those of the publisher, the editors and the
the data presented, we project that, when used intentionally, it reviewers. Any product that may be evaluated in this article, or

Frontiers in Immunology 11 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

claim that may be made by its manufacturer, is not guaranteed or now the imputed data is visualized. (B) Same scatter plot as in Figure 2D,
where now the imputed data is used. (C) UMAP density plots highlighting the
endorsed by the publisher.
cells belonging to the categories created in (B). The parameter “vmin” to
control the lower limit in the color scale was set to 0.05. (D) The same UMAP
plot and selection as in Figure 3A visualizing the imputed data. The parameter
“vmax” to control the upper limit in the color scale was set to 0.12.
Supplementary material
SUPPLEMENTARY FIGURE 2
The Supplementary Material for this article can be found online at:
(A) Guassian mixture model of RORC and TBX21 mRNA-derived UMI count
https://www.frontiersin.org/articles/10.3389/fimmu.2023.1027346/ with the data from Figure 3, plotted on the same UMAP plot. (B) The same plot
full#supplementary-material as Figure 5A, with the modification that the CD8A (left), CD8B (middle) and
CD4 (right) mRNA derived UMI count is highlighted as indicated. (C) UMAP
SUPPLEMENTARY FIGURE 1 plot with CD8 (right) and CD4 (left) antibody UMI counts highlighted in red.
Visualization of how the previously set selections in Figures 2, 3 and look with (D) Guassian mixture model of CD4 and CD8A mRNA-derived UMI count with
imputed data. (A) The same UMAP plot and selection as in Figure 2D, where the data from Figure 5B, plotted on the same UMAP plot.

References
1. Schett G, McInnes IB, Neurath MF. Reframing immune-mediated inflammatory 20. Cellxgene Data Portal. Cellxgene data portal. Available at: https://cellxgene.
diseases through signature cytokine hubs. N Engl J Med (2021) 385:628–39. cziscience.com/ (Accessed 8, 2022).
doi: 10.1056/NEJMra1909094 21. Hunter JD. Matplotlib: a 2D graphics environment. Comput Sci Eng (2007)
2. Roda G, Chien Ng S, Kotze PG, Argollo M, Panaccione R, Spinelli A, et al. 9:90–5. doi: 10.1109/MCSE.2007.55
Crohn’s disease. Nat Rev Dis Primer (2020) 6:22. doi: 10.1038/s41572-020-0156-2 22. Harris CR, Millman KJ, van der Walt SJ, Gommers R, Virtanen P, Cournapeau
3. Danese S, Fiocchi C. Ulcerative colitis. N Engl J Med (2011) 365:1713–25. D, et al. Array programming with NumPy. Nature (2020) 585:357–62. doi: 10.1038/
doi: 10.1056/NEJMra1102942 s41586-020-2649-2
4. Fuss IJ, Neurath M, Boirivant M, Klein JS, de la Motte C, Strong SA, et al. 23. McInnes L, Healy J, Saul N, Großberger L. UMAP: uniform manifold
Disparate CD4+ lamina propria (LP) lymphokine secretion profiles in inflammatory approximation and projection. J Open Source Softw (2018) 3:861. doi: 10.21105/
bowel disease. crohn’s disease LP cells manifest increased secretion of IFN-gamma, joss.00861
whereas ulcerative colitis LP cells manifest increased secretion of IL-5. J Immunol 24. Maaten van der L. Accelerating t-SNE using tree-based algorithms. J Mach
Baltim Md 1950 (1996) 157:1261–70. Learn Res (2014) 15:3221–45.
5. Jostins L, Ripke S, Weersma RK, Duerr RH, McGovern DP, Hui KY, et al. Host- 25. Virshup I, Rybakov S, Theis FJ, Angerer P, Wolf FA. Anndata: annotated data.
microbe interactions have shaped the genetic architecture of inflammatory bowel BioRxiv (2021) 2021:12.16.473007. doi: 10.1101/2021.12.16.473007
disease. Nature (2012) 491:119–24. doi: 10.1038/nature11582
26. Becker E, Dedden M, Gall C, Wiendl M, Ekici AB, Schulz-Kuhnt A, et al.
6. Chang JT. Pathophysiology of inflammatory bowel diseases. N Engl J Med (2020) Residual homing of a4b7-expressing b1+PI16+ regulatory T cells with potent
383:2652–64. doi: 10.1056/NEJMra2002697 suppressive activity correlates with exposure-efficacy of vedolizumab. Gut (2022)
7. Neurath MF. Targeting immune cell circuits and trafficking in inflammatory 71:1551–66. doi: 10.1136/gutjnl-2021-324868
bowel disease. Nat Immunol (2019) 20:970–9. doi: 10.1038/s41590-019-0415-0 27. van Dijk D, Sharma R, Nainys J, Yim K, Kathail P, Carr AJ, et al. Recovering
8. Zundler S, Becker E, Schulze LL, Neurath MF. Immune cell trafficking and gene interactions from single-cell data using data diffusion. Cell (2018) 174:716–
retention in inflammatory bowel disease: mechanistic insights and therapeutic 729.e27. doi: 10.1016/j.cell.2018.05.061
advances. Gut (2019) 68:1688–700. doi: 10.1136/gutjnl-2018-317977 28. Sturm G, Szabo T, Fotakis G, Haider M, Rieder D, Trajanoski Z, et al. Scirpy: a
9. Neurath MF. Cytokines in inflammatory bowel disease. Nat Rev Immunol (2014) scanpy extension for analyzing single-cell T-cell receptor-sequencing data.
14:329–42. doi: 10.1038/nri3661 Bioinformatics (2020) 36:4817–8. doi: 10.1093/bioinformatics/btaa611
10. Maul J, Loddenkemper C, Mundt P, Berg E, Giese T, Stallmach A, et al. 29. Park CO, Kupper TS. The emerging role of resident memory T cells in protective
Peripheral and intestinal regulatory CD4+ CD25(high) T cells in inflammatory immunity and inflammatory disease. Nat Med (2015) 21:688–97. doi: 10.1038/nm.3883
bowel disease. Gastroenterology (2005) 128:1868–78. doi: 10.1053/j.gastro.2005.03.043 30. Neurath MF, Weigmann B, Finotto S, Glickman J, Nieuwenhuis E, Iijima H,
11. Feagan BG, Sandborn WJ, Gasink C, Jacobstein D, Lang Y, Friedman JR, et al. et al. The transcription factor T-bet regulates mucosal T cell activation in
Ustekinumab as induction and maintenance therapy for crohn’s disease. N Engl J Med experimental colitis and crohn’s disease. J Exp Med (2002) 195:1129–43.
(2016) 375:1946–60. doi: 10.1056/NEJMoa1602773 doi: 10.1084/jem.20011956
12. Sandborn WJ, Feagan BG, D’Haens G, Wolf DC, Jovanovic I, Hanauer SB, et al. 31. Leppkes M, Becker C, Ivanov II, Hirth S, Wirtz S, Neufert C, et al. RORgamma-
Ozanimod as induction and maintenance therapy for ulcerative colitis. N Engl J Med expressing Th17 cells induce murine chronic intestinal inflammation via redundant
(2021) 385:1280–91. doi: 10.1056/NEJMoa2033617 effects of IL-17A and IL-17F. Gastroenterology (2009) 136:257–67. doi: 10.1053/
13. Feagan BG, Danese S, Loftus EV, Vermeire S, Schreiber S, Ritter T, et al. j.gastro.2008.10.018
Filgotinib as induction and maintenance therapy for ulcerative colitis (SELECTION): a 32. Annunziato F, Cosmi L, Santarlasci V, Maggi L, Liotta F, Mazzinghi B, et al.
phase 2b/3 double-blind, randomised, placebo-controlled trial. Lancet (2021) Phenotypic and functional features of human Th17 cells. J Exp Med (2007) 204:1849–
397:2372–84. doi: 10.1016/S0140-6736(21)00666-8 61. doi: 10.1084/jem.20070663
14. Hu X, Zhou X. Impact of single-cell RNA sequencing on understanding immune 33. Gosselin A, Monteiro P, Chomont N, Diaz-Griffero F, Said EA, Fonseca S, et al.
regulation. J Cell Mol Med (2022) 26:4645–57. doi: 10.1111/jcmm.17493 Peripheral blood CCR4+CCR6+ and CXCR3+CCR6+ CD4+ T cells are highly
15. Tang F, Barbacioru C, Wang Y, Nordman E, Lee C, Xu N, et al. mRNA-seq permissive to HIV-1 infection. J Immunol (2010) 184:1604–16. doi: 10.4049/
whole-transcriptome analysis of a single cell. Nat Methods (2009) 6:377–82. jimmunol.0903058
doi: 10.1038/nmeth.1315 34. Sandborn WJ, Feagan BG, Rutgeerts P, Hanauer S, Colombel J-F, Sands BE, et al.
16. Andrews TS, Kiselev VY, McCarthy D, Hemberg M. Tutorial: guidelines for the Vedolizumab as induction and maintenance therapy for crohn’s disease. N Engl J Med
computational analysis of single-cell RNA sequencing data. Nat Protoc (2021) 16:1–9. (2013) 369:711–21. doi: 10.1056/NEJMoa1215739
doi: 10.1038/s41596-020-00409-w 35. Feagan BG, Rutgeerts P, Sands BE, Hanauer S, Colombel J-F, Sandborn WJ, et al.
17. Wolf FA, Angerer P, Theis FJ. SCANPY: large-scale single-cell gene expression Vedolizumab as induction and maintenance therapy for ulcerative colitis. N Engl J Med
data analysis. Genome Biol (2018) 19:15. doi: 10.1186/s13059-017-1382-0 (2013) 369:699–710. doi: 10.1056/NEJMoa1215734

18. Hao Y, Hao S, Andersen-Nissen E, Mauck WM, Zheng S, Butler A, et al. 36. Zhu H, Galdos FX, Lee D, Waliany S, Huang YV, Ryan J, et al. Identification of
Integrated analysis of multimodal single-cell data. Cell (2021) 184:3573–3587.e29. pathogenic immune cell subsets associated with checkpoint inhibitor-induced
doi: 10.1016/j.cell.2021.04.048 myocarditis. Circulation (2022) 146:316–35. doi: 10.1161/CIRCULATIONAHA.
121.056730
19. Speir ML, Bhaduri A, Markov NS, Moreno P, Nowakowski TJ, Papatheodorou I,
et al. UCSC cell browser: visualize your single-cell data. Bioinforma Oxf Engl (2021) 37. McKinnon KM. Flow cytometry: an overview. Curr Protoc Immunol (2018)
37:4578–80. doi: 10.1093/bioinformatics/btab503 120:5.1.1–5.1.11. doi: 10.1002/cpim.40

Frontiers in Immunology 12 frontiersin.org

Dedden et al. 10.3389/fimmu.2023.1027346

38. Adan A, Alizada G, Kiraz Y, Baran Y, Nalbant A. Flow cytometry: basic 40. Jagla B, Libri V, Chica C, Rouilly V, Mella S, Puceat M, et al. SCHNAPPs - single cell
principles and applications. Crit Rev Biotechnol (2017) 37:163–76. doi: 10.3109/ sHiNy APPlication(s). J Immunol Methods (2021) 499:113176. doi: 10.1016/j.jim.2021.113176
07388551.2015.1128876 41. Traag VA, Waltman L, van Eck NJ. From louvain to Leiden: guaranteeing well-
39. Zhang MJ, Ntranos V, Tse D. Determining sequencing depth in a single- connected communities. Sci Rep (2019) 9:5233. doi: 10.1038/s41598-019-41695-z
cell RNA-seq experiment. Nat Commun (2020) 11:774. doi: 10.1038/s41467- 42. Skinnider MA, Squair JW, Courtine G. Enabling reproducible re-analysis of
020-14482-y single-cell data. Genome Biol (2021) 22:215. doi: 10.1186/s13059-021-02422-y

Frontiers in Immunology 13 frontiersin.org