| Title: | Data, Clustering Results and Visualization Functions From Duò et al (2018) |
|---|---|
| Description: | Preprocessed experimental and simulated scRNA-seq data sets used for evaluation of clustering methods for scRNA-seq data in Duò et al (2018). Also contains results from applying several clustering methods to each of the data sets, and functions for plotting method performance. |
| Authors: | Angelo Duò, Charlotte Soneson |
| Maintainer: | Angelo Duò <[email protected]> |
| License: | GPL (>=2) |
| Version: | 1.30.0 |
| Built: | 2026-10-10 11:13:50 UTC |
| Source: | https://github.com/bioc/DuoClustering2018 |
Help function for computing ARI
ari_df(x)ari_df(x)
x |
A |
a data.frame with ARI values for each pair of runs.
Clustering results for the performance evaluation of clustering methods for scRNA-seq data, corresponding to v1 of Duò et al. (2018).
clustering_summary_filteredExpr10_Koh_v1(metadata = FALSE)clustering_summary_filteredExpr10_Koh_v1(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
These objects contain clustering results from the performance evaluation of
clustering methods for scRNA-seq data. The clustering results are provided as
a data.frame object containing 10 variables (columns) named
dataset, method, cell, run, k,
resolution, cluster, trueclass, est_k and
elapsed. For further information see Duò et al. (2018).
Returns a data.frame.
Duò A, Robinson MD. and Soneson C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res., 7:1141.
clustering_summary_filteredExpr10_Koh_v1()clustering_summary_filteredExpr10_Koh_v1()
Clustering results for the performance evaluation of clustering methods for scRNA-seq data, corresponding to v2 of Duò et al. (2018).
clustering_summary_filteredExpr10_Koh_v2(metadata = FALSE)clustering_summary_filteredExpr10_Koh_v2(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
These objects contain clustering results from the performance evaluation of
clustering methods for scRNA-seq data. The clustering results are provided as
a data.frame object containing 10 variables (columns) named
dataset, method, cell, run, k,
resolution, cluster, trueclass, est_k and
elapsed. For further information see Duò et al. (2018).
Returns a data.frame.
Duò A, Robinson MD. and Soneson C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res., 7:1141.
clustering_summary_filteredExpr10_Koh_v2()clustering_summary_filteredExpr10_Koh_v2()
Hyperparameter values for all clustering algorithms and data sets in v1 of Duo et al (F1000Research 2018)
duo_clustering_all_parameter_settings_v1(metadata = FALSE)duo_clustering_all_parameter_settings_v1(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
List of hyperparameter values used for all clustering algorithms and data sets in v1 of Duò et al (F1000Research 2018).
Returns a list with hyperparameter values for all data sets and
methods.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Research, 7:1141.
duo_clustering_all_parameter_settings_v1()duo_clustering_all_parameter_settings_v1()
Hyperparameter values for all clustering algorithms and data sets in v2 of Duo et al (F1000Research 2018)
duo_clustering_all_parameter_settings_v2(metadata = FALSE)duo_clustering_all_parameter_settings_v2(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
List of hyperparameter values used for all clustering algorithms and data sets in v2 of Duò et al (F1000Research 2018).
Returns a list with hyperparameter values for all data sets and
methods.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Research, 7:1141.
duo_clustering_all_parameter_settings_v2()duo_clustering_all_parameter_settings_v2()
Data package containing scRNA-seq data sets, clustering results and functions for summarizing the performance of different scRNA-seq clustering methods.
This package contains publicly available scRNA-seq data sets and the
accompanying results from clustering using general-purpose methods and scRNA-seq
clustering methods. Several real data sets a well as simulated data sets are
provided. The data sets have been used to evaluate the performance of clustering
algorithms in our previous work and publication (Duò et al.,
F1000Research 2018). The data sets are available as
SingleCellExperiment objects. For additional details on the data
sets, see the help files for the respective data sets.
Additionally, the clustering results from the evaluation as well as functions for summarization and visualization of the clustering results are provided.
A description of the basic usage of the package for retrieving data sets and clustering results, and how to construct various plots summarizing the performance of different methods is outlined in the package vignettes.
Angelo Duò and Charlotte Sonesson
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Research, 7:1141.
Plot entropy of cluster assignments
plot_entropy(res, method_colors = NULL)plot_entropy(res, method_colors = NULL)
res |
A data.frame with clustering results. |
method_colors |
A named vector with colors to use for the different clustering methods. Can be NULL, in which case colors are chosen automatically. |
A named list of ggplot2 objects
Angelo Duo, Charlotte Soneson
res <- clustering_summary_filteredExpr10_Koh_v2() plots <- plot_entropy(res)res <- clustering_summary_filteredExpr10_Koh_v2() plots <- plot_entropy(res)
Plot differences between optimal, estimated and true number of clusters
plot_k_diff(res, method_colors = NULL)plot_k_diff(res, method_colors = NULL)
res |
A data.frame with clustering results. |
method_colors |
A named vector with colors to use for the different clustering methods. Can be NULL, in which case colors are chosen automatically. |
A named list of ggplot2 objects
Angelo Duo, Charlotte Soneson
res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_k_diff(res)res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_k_diff(res)
Generate various plots of the agreement between each clustering and the true partitioning of the cells, quantified by the adjusted Rand index (ARI).
plot_performance(res, method_colors = NULL)plot_performance(res, method_colors = NULL)
res |
A data.frame with clustering results. |
method_colors |
A named vector with colors to use for the different clustering methods. Can be NULL, in which case colors are chosen automatically. |
A named list of ggplot2 objects
Angelo Duo, Charlotte Soneson
res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_performance(res)res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_performance(res)
Plot the stability of the clusterings obtained for each method
plot_stability(res, method_colors = NULL)plot_stability(res, method_colors = NULL)
res |
A data.frame with clustering results. |
method_colors |
A named vector with colors to use for the different clustering methods. Can be NULL, in which case colors are chosen automatically. |
A named list of ggplot2 objects
Angelo Duo, Charlotte Soneson
res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_stability(res)res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_stability(res)
Plot the elapsed time for each clustering method
plot_timing(res, method_colors = NULL, scaleMethod = NULL)plot_timing(res, method_colors = NULL, scaleMethod = NULL)
res |
A data.frame with clustering results. |
method_colors |
A named vector with colors to use for the different clustering methods. Can be NULL, in which case colors are chosen automatically. |
scaleMethod |
Either NULL or one of the clustering methods in the result
data.frame. If not NULL, a plot will be generated where all elapsed times
are normalized by dividing with the time for |
A named list of ggplot2 objects
Angelo Duo, Charlotte Soneson
res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_timing(res)res <- clustering_summary_filteredExpr10_Koh_v1() plots <- plot_timing(res)
Gene or TCC counts for a scRNA-seq data set from Koh et al. (2016), consisting of in vitro cultured H7 embryonic stem cells (WiCell) and H7-derived downstream early mesoderm progenitors.
sce_full_Koh(metadata = FALSE) sce_filteredExpr10_Koh(metadata = FALSE) sce_filteredHVG10_Koh(metadata = FALSE) sce_filteredM3Drop10_Koh(metadata = FALSE) sce_full_KohTCC(metadata = FALSE) sce_filteredExpr10_KohTCC(metadata = FALSE) sce_filteredHVG10_KohTCC(metadata = FALSE) sce_filteredM3Drop10_KohTCC(metadata = FALSE)sce_full_Koh(metadata = FALSE) sce_filteredExpr10_Koh(metadata = FALSE) sce_filteredHVG10_Koh(metadata = FALSE) sce_filteredM3Drop10_Koh(metadata = FALSE) sce_full_KohTCC(metadata = FALSE) sce_filteredExpr10_KohTCC(metadata = FALSE) sce_filteredHVG10_KohTCC(metadata = FALSE) sce_filteredM3Drop10_KohTCC(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
SingleCellExperiment
This is a scRNA-seq data set originally from Koh et al. (2016). The data set consists of gene-level read counts or TCCs (transcript compatibility counts) of in vitro cultured human H7 embryonic stem cells (WiCell) and H7-derived downstream early mesoderm progenitors. It contains 9 subpopulations, defined by the cell phenotype given by the authors' annotations. The data sets have been used to evaluate the performance of clustering algorithms in Duò et al. (2018).
For the sce_full_Koh data set, all genes except those with zero counts
across all cells are retained. The gene counts are
gene-level length-scaled TPM values derived from Salmon (Patro et al. (2017))
quantifications (see
Soneson and Robinson (2018)). For the TCC data set we estimated transcripts
compatibility counts using kallisto as an alternative to the gene-level
count matrix (Bray et al. (2016), Ntranos et al. (2016)).
The scater package was used to perform quality control of the data sets
(McCarthy et al. (2017)).
Features with zero counts across all cells, as well as all cells with total
count or total number
of detected features more than 3 median absolute deviations (MADs) below the
median across all cells (on the log scale), were excluded.
The sce_full_Koh data set consists of 531 cells and 48,981 features,
and the sce_full_KohTCC data set of 531 cells and 811,938 features.
The filteredExpr, filteredHVG and filteredM3Drop10 are
further reduced data sets.
For each of the filtering methods, we retained 10 percent of the number of genes
(with a non-zero count in at least one cell) in the original data sets.
For the filteredExpr data sets, only the genes/TCCs with the highest
average expression (log-normalized count) value across all cells were retained.
Using the Seurat package (Satija et al. (2015)), the filteredHVG
data sets were filtered on the variability of the features and only the most
highly variable ones were retained. Finally, the M3Drop package was
used to model the dropout rate of the genes as a function of the mean
expression level using the Michaelis-Menten equation and select variables to
retain for the filteredM3Drop10 data sets (Andrews and Hemberg (2018)).
The scater package was used to normalize the count values, based on
normalization factors calculated by the deconvolution method from the
scran package (Lun et al. (2016)).
This data set is provided as a SingleCellExperiment object
(Lun and Risso (2017)). Raw data files for the original data set (SRP073808)
are available from https://www.ncbi.nlm.nih.gov/sra?term=SRP073808.
Returns a SingleCellExperiment object.
Andrews, T.S., and Hemberg, M. (2018). Dropout-based feature selection for scRNASeq. bioRxiv doi:https://doi.org/10.1101/065094.
Bray, N.L., Pimentel, H., Melsted, P., and Pachter, L. (2016). Near-optimal probabilistic RNA-seq quantification. Nat. Biotechnol. 34: 525–527.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res. 7:1141.
Koh, P.W., Sinha, R., Barkal, A.A., Morganti R.M., Chen, A., Weissman, I.L., Ang, L.T., Kundaje, A., and Loh, K.M. (2016). An atlas of transcriptional, chromatin accessibility, and surface marker changes in human mesoderm development. Scientific Data 3:160109.
Lun, A.T.L., Bach, K., and Marioni, J.C. (2016) Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 17(1): 75.
Lun, A.T.L., and Risso, D. (2017). SingleCellExperiment: S4 Classes for Single Cell Data. R package version 1.0.0.
McCarthy, D.J., Campbell, K.R., Lun, A.T.L., and Wills, Q.F. (2017): Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33(8): 1179-1186.
Ntranos, V., Kamath, G.M., Zhang, J.M., Pachter, L., and Tse, D.N. (2016): Fast and accurate single-cell RNA-seq analysis by clustering of transcript-compatibility counts. Genome Biol. 17:112.
Patro, R., Duggal, G., Love, M.I., Irizarry, R.A., and Kingsford, C. (2017): Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14:417-419.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33(5): 495–502.
Soneson, C., and Robinson, M.D. (2018). Bias, robustness and scalability in single-cell differential expression analysis. Nat. Methods, 15(4): 255-261.
sce_filteredHVG10_Koh()sce_filteredHVG10_Koh()
Gene or TCC counts for scRNA-seq data set from Kumar et al. (2014), consisting of mESCs with various genetic perturbations which are cultured in different media.
sce_full_Kumar(metadata = FALSE) sce_filteredExpr10_Kumar(metadata = FALSE) sce_filteredHVG10_Kumar(metadata = FALSE) sce_filteredM3Drop10_Kumar(metadata = FALSE) sce_full_KumarTCC(metadata = FALSE) sce_filteredExpr10_KumarTCC(metadata = FALSE) sce_filteredHVG10_KumarTCC(metadata = FALSE) sce_filteredM3Drop10_KumarTCC(metadata = FALSE)sce_full_Kumar(metadata = FALSE) sce_filteredExpr10_Kumar(metadata = FALSE) sce_filteredHVG10_Kumar(metadata = FALSE) sce_filteredM3Drop10_Kumar(metadata = FALSE) sce_full_KumarTCC(metadata = FALSE) sce_filteredExpr10_KumarTCC(metadata = FALSE) sce_filteredHVG10_KumarTCC(metadata = FALSE) sce_filteredM3Drop10_KumarTCC(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
SingleCellExperiment
This is a scRNA-seq data set originally from Kumar et al. (2014). The data set consists of gene-level read counts or TCCs (transcript compatibility counts) for mESCs from Mus musculus with various genetic perturbations which are cultured in different media. It contains 3 subpopulations, defined by the cell phenotype given by the authors' annotations. The data sets have been used to evaluate the performance of clustering algorithms in Duò et al. (2018).
For the sce_full_Kumar data set, all genes except those with zero counts
across all cells are retained. The gene counts are
gene-level length-scaled TPM values derived from Salmon (Patro et al. (2017))
quantifications (see
Soneson and Robinson (2018)). For the TCC data set we estimated transcripts
compatibility counts using kallisto as an alternative to the gene-level
count matrix (Bray et al. (2016), Ntranos et al. (2016)).
The scater package was used to perform quality control of the data sets
(McCarthy et al. (2017)).
Features with zero counts across all cells, as well as all cells with total
count or total number of detected features more than 3 median absolute
deviations (MADs) below the median across all cells (on the log scale), were
excluded. Additionally, cells with a large fraction of ERCC reads were
filtered out.
The sce_full_Kumar data set consists of 246 cells and 45,159 features,
the sce_full_KumarTCC data set of 246 cells and 803,405 features,
respectively.
The filteredExpr, filteredHVG and filteredM3Drop10 are
further reduced data sets.
For each filtering method, we retained 10 percent of the original number
of genes
(with a non-zero count in at least one cell) in the original data sets.
For the filteredExpr data sets, only the genes/TCCs with the highest
average expression (log-normalized count) value across all cells were retained.
Using the Seurat package, the filteredHVG data sets were filtered
on the variability of the features and only the most highly variable ones were
retained (Satija et al. (2015)). Finally, the M3Drop package was used
to model the dropout rate of the genes as a function of the mean expression
level using the Michaelis-Menten equation and select variables to retain for
the filteredM3Drop10 data sets (Andrews and Hemberg (2018)).
The scater package was used to normalize the count values, based on
normalization factors calculated by the deconvolution method from the
scran package (Lun et al. (2016)).
This data set is provided as a SingleCellExperiment object
(Lun and Risso (2017)). For further information on the
SingleCellExperiment class, see the corresponding manual.
Raw data files for the original data set (GSE60749) are available from
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE60749.
Returns a SingleCellExperiment object.
Andrews, T.S., and Hemberg, M. (2018). Dropout-based feature selection for scRNASeq. bioRxiv doi:https://doi.org/10.1101/065094.
Bray, N.L., Pimentel, H., Melsted, P., and Pachter, L. (2016). Near-optimal probabilistic RNA-seq quantification. Nat. Biotechnol. 34: 525–527.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res. 7:1141.
Kumar R.M., Cahan P., Shalek A.K., Satija R., DaleyKeyser A.J., Li H., Zhang J., Pardee K., Gennert D., Trombetta J.J., Ferrante T.C., Regev A., Daley G.Q., and Collins J.J. (2014) Deconstructing transcriptional heterogeneity in pluripotent stem cells. Nature 516(7529): 56–61.
Lun, A.T.L., Bach, K., and Marioni, J.C. (2016) Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 17(1): 75.
Lun, A.T.L., and Risso, D. (2017). SingleCellExperiment: S4 Classes for Single Cell Data. R package version 1.0.0.
McCarthy, D.J., Campbell, K.R., Lun, A.T.L., and Wills, Q.F. (2017): Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33(8): 1179-1186.
Ntranos, V., Kamath, G.M., Zhang, J.M., Pachter, L., and Tse, D.N. (2016): Fast and accurate single-cell RNA-seq analysis by clustering of transcript-compatibility counts. Genome Biol. 17:112.
Patro, R., Duggal, G., Love, M.I., Irizarry, R.A., and Kingsford, C. (2017): Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14:417-419.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33(5): 495–502.
Soneson, C., and Robinson, M.D. (2018). Bias, robustness and scalability in single-cell differential expression analysis. Nat. Methods 15(4): 255-261.
sce_filteredExpr10_Kumar()sce_filteredExpr10_Kumar()
Gene counts for scRNA-seq data sets simulated with the splatter package.
sce_full_SimKumar4easy(metadata = FALSE) sce_filteredExpr10_SimKumar4easy(metadata = FALSE) sce_filteredHVG10_SimKumar4easy(metadata = FALSE) sce_filteredM3Drop10_SimKumar4easy(metadata = FALSE) sce_full_SimKumar4hard(metadata = FALSE) sce_filteredExpr10_SimKumar4hard(metadata = FALSE) sce_filteredHVG10_SimKumar4hard(metadata = FALSE) sce_filteredM3Drop10_SimKumar4hard(metadata = FALSE) sce_full_SimKumar8hard(metadata = FALSE) sce_filteredExpr10_SimKumar8hard(metadata = FALSE) sce_filteredHVG10_SimKumar8hard(metadata = FALSE) sce_filteredM3Drop10_SimKumar8hard(metadata = FALSE)sce_full_SimKumar4easy(metadata = FALSE) sce_filteredExpr10_SimKumar4easy(metadata = FALSE) sce_filteredHVG10_SimKumar4easy(metadata = FALSE) sce_filteredM3Drop10_SimKumar4easy(metadata = FALSE) sce_full_SimKumar4hard(metadata = FALSE) sce_filteredExpr10_SimKumar4hard(metadata = FALSE) sce_filteredHVG10_SimKumar4hard(metadata = FALSE) sce_filteredM3Drop10_SimKumar4hard(metadata = FALSE) sce_full_SimKumar8hard(metadata = FALSE) sce_filteredExpr10_SimKumar8hard(metadata = FALSE) sce_filteredHVG10_SimKumar8hard(metadata = FALSE) sce_filteredM3Drop10_SimKumar8hard(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
SingleCellExperiment
Using one subpopulation of the sce_full_Kumar data set as input,
scRNA-seq data with known group structure was simulated with the splatter
package from Zappia et al. (2017). The simulated data have been used to evaluate
the performance of clustering algorithms in Duò et al. (2018).
Three data sets have been generated, each consisting of 500 cells and
approximately 43,000 features, with varying degree of cluster separability.
The sce_full_SimKumar4easy data set consists of 4 subpopulations with
relative abundances 0.1, 0.15, 0.5 and 0.25, and probabilities of differential
expression set to 0.05, 0.1, 0.2 and 0.4 for the four subpopulations,
respectively. The sce_full_SimKumar4hard data set consists of 4
subpopulations with relative abundances 0.2, 0.15, 0.4 and 0.25, and
probabilities of differential expression 0.01, 0.05, 0.05 and 0.08. Finally,
the sce_full_SimKumar8hard data set consists of 8 subpopulations with
relative abundances 0.13, 0.07, 0.1, 0.05, 0.4, 0.1, 0.1 and 0.05, and
probabilities of differential expression equal to 0.03, 0.03, 0.03, 0.05, 0.05,
0.07, 0.08 and 0.1, respectively.
The scater package was used to perform quality control of the data sets
(McCarthy et al. (2017)).
Features with zero counts across all cells, as well as cells with total count
or total number of detected features more than 3 median absolute deviations
(MADs) below the median across all cells (on the log scale), were excluded.
The filteredExpr, filteredHVG and filteredM3Drop10 are
further reduced data sets. For each of the filtering method, we retained 10
percent of the original number of genes
(with a non-zero count in at least one cell) in the original data sets.
For the filteredExpr data sets, only the genes with the highest average
expression (log-normalized count) value across all cells were retained.
Using the Seurat package, the filteredHVG data sets were filtered
on the variability of the features and only the most highly variable ones were
retained (Satija et al. (2015)). Finally, the M3Drop package was used
to model the dropout rate of the genes as a function of the mean expression
level using the Michaelis-Menten equation and select variables to retain for
the filteredM3Drop10 data sets (Andrews and Hemberg (2018)).
The scater package was used to normalize the count values, based on
normalization factors calculated by the deconvolution method from the
scran package (Lun et al. (2016)).
This data set is provided as a SingleCellExperiment object
(Lun and Risso (2017)). For further information on the
SingleCellExperiment class, see the corresponding manual.
Returns a SingleCellExperiment object.
Andrews, T.S., and Hemberg, M. (2018). Dropout-based feature selection for scRNASeq. bioRxiv doi:https://doi.org/10.1101/065094.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res. 7:1141.
Lun, A.T.L., Bach, K., and Marioni, J.C. (2016) Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 17(1): 75.
Lun, A.T.L., and Risso, D. (2017). SingleCellExperiment: S4 Classes for Single Cell Data. R package version 1.0.0.
McCarthy, D.J., Campbell, K.R., Lun, A.T.L., and Wills, Q.F. (2017): Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33(8): 1179-1186.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33(5): 495–502.
Zappia, L., Phipson, B., and Oshlack, A. (2017). Splatter: simulation of single-cell RNA sequencing data. Genome Biol. 18(1): 174.
sce_filteredExpr10_SimKumar4easy()sce_filteredExpr10_SimKumar4easy()
Gene or TCC counts for scRNA-seq data set from Trapnell et al. (2014), consisting of primary myoblasts over a time course of serum-induced differentiation.
sce_full_Trapnell(metadata = FALSE) sce_filteredExpr10_Trapnell(metadata = FALSE) sce_filteredHVG10_Trapnell(metadata = FALSE) sce_filteredM3Drop10_Trapnell(metadata = FALSE) sce_full_TrapnellTCC(metadata = FALSE) sce_filteredExpr10_TrapnellTCC(metadata = FALSE) sce_filteredHVG10_TrapnellTCC(metadata = FALSE) sce_filteredM3Drop10_TrapnellTCC(metadata = FALSE)sce_full_Trapnell(metadata = FALSE) sce_filteredExpr10_Trapnell(metadata = FALSE) sce_filteredHVG10_Trapnell(metadata = FALSE) sce_filteredM3Drop10_Trapnell(metadata = FALSE) sce_full_TrapnellTCC(metadata = FALSE) sce_filteredExpr10_TrapnellTCC(metadata = FALSE) sce_filteredHVG10_TrapnellTCC(metadata = FALSE) sce_filteredM3Drop10_TrapnellTCC(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
SingleCellExperiment
This is a scRNA-seq data set originally from Trapnell et al. (2014). The data set consists of gene-level read counts or TCCs (transcript compatibility counts) from human primary myoblasts over a time course of serum-induced differentiation. It contains 3 subpopulations, defined by the cell phenotype given by the authors' annotations. The data sets have been used to evaluate the performance of clustering algorithms in Duò et al. (2018).
For the sce_full_Trapnell data set, all genes except those with zero
counts across all cells are retained. The gene counts are
gene-level length-scaled TPM values derived from Salmon (Patro et al. (2017))
quantifications (see
Soneson and Robinson (2018)). For the TCC data set we estimated transcripts
compatibility counts using kallisto as an alternative to the gene-level
count matrix (Bray et al. (2016), Ntranos et al. (2016)).
The scater package was used to perform quality control of the data sets
(McCarthy et al. (2017)).
Features with zero counts across all cells, as well as all cells with total
count or total number of detected features more than 3 median absolute
deviations (MADs) below the median across all cells (on the log scale),
were excluded. Additionally, cells that were classified as doublets or
debris were filtered out.
The sce_full_Trapnell data set consists of 222 cells and 41,111 features,
the sce_full_TrapnellTCC data set of 227 cells and 684,953 features,
respectively.
The filteredExpr, filteredHVG and filteredM3Drop10 are
further reduced data sets.
For each of the filtering method, we retained 10 percent of the original
number of genes
(with a non-zero count in at least one cell) in the original data sets.
For the filteredExpr data sets, only the genes/TCCs with the highest
average expression (log-normalized count) value across all cells were retained.
Using the Seurat package, the filteredHVG data sets were filtered
on the variability of the features and only the most highly variable ones were
retained (Satija et al. (2015)). Finally, the M3Drop package was used
to model the dropout rate of the genes as a function of the mean expression
level using the Michaelis-Menten equation and select variables to retain for
the filteredM3Drop10 data sets (Andrews and Hemberg (2018)).
The scater package was used to normalize the count values, based on
normalization factors calculated by the deconvolution method from the
scran package (Lun et al. (2016)).
This data set is provided as a SingleCellExperiment object
(Lun and Risso (2017)). For further information on the
SingleCellExperiment class, see the corresponding manual.
Raw data files for the original data set (GSE52529) are available from
https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE52529.
Returns a SingleCellExperiment object.
Andrews, T.S., and Hemberg, M. (2018). Dropout-based feature selection for scRNASeq. bioRxiv doi:https://doi.org/10.1101/065094.
Bray, N.L., Pimentel, H., Melsted, P., and Pachter, L. (2016). Near-optimal probabilistic RNA-seq quantification. Nat. Biotechnol. 34: 525–527.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res. 7:1141.
Lun, A.T.L., Bach, K., and Marioni, J.C. (2016) Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 17(1): 75.
Lun, A.T.L., and Risso, D. (2017). SingleCellExperiment: S4 Classes for Single Cell Data. R package version 1.0.0.
McCarthy, D.J., Campbell, K.R., Lun, A.T.L., and Wills, Q.F. (2017): Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33(8): 1179-1186.
Ntranos, V., Kamath, G.M., Zhang, J.M., Pachter, L., and Tse, D.N. (2016): Fast and accurate single-cell RNA-seq analysis by clustering of transcript-compatibility counts. Genome Biol. 17:112.
Patro, R., Duggal, G., Love, M.I., Irizarry, R.A., and Kingsford, C. (2017): Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14:417-419.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33(5): 495–502.
Soneson, C., and Robinson, M.D. (2018). Bias, robustness and scalability in single-cell differential expression analysis. Nat. Methods, 15(4): 255-261.
Trapnell, C., Cacchiarelli, D., Grimsby, J., Pokharel, P., Li, S., Morse, M., Lennon, N.J., Livak, K.J., Mikkelsen, T.S., and Rinn, J.L. (2014). The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32(4): 381–386.
sce_filteredExpr10_Trapnell()sce_filteredExpr10_Trapnell()
Gene counts for scRNA-seq data sets from Zheng et al. (2017), consisting of pre-sorted cell types combined into three artificial data sets with different cell proportions.
sce_full_Zhengmix4eq(metadata = FALSE) sce_filteredExpr10_Zhengmix4eq(metadata = FALSE) sce_filteredHVG10_Zhengmix4eq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix4eq(metadata = FALSE) sce_full_Zhengmix4uneq(metadata = FALSE) sce_filteredExpr10_Zhengmix4uneq(metadata = FALSE) sce_filteredHVG10_Zhengmix4uneq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix4uneq(metadata = FALSE) sce_full_Zhengmix8eq(metadata = FALSE) sce_filteredExpr10_Zhengmix8eq(metadata = FALSE) sce_filteredHVG10_Zhengmix8eq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix8eq(metadata = FALSE)sce_full_Zhengmix4eq(metadata = FALSE) sce_filteredExpr10_Zhengmix4eq(metadata = FALSE) sce_filteredHVG10_Zhengmix4eq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix4eq(metadata = FALSE) sce_full_Zhengmix4uneq(metadata = FALSE) sce_filteredExpr10_Zhengmix4uneq(metadata = FALSE) sce_filteredHVG10_Zhengmix4uneq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix4uneq(metadata = FALSE) sce_full_Zhengmix8eq(metadata = FALSE) sce_filteredExpr10_Zhengmix8eq(metadata = FALSE) sce_filteredHVG10_Zhengmix8eq(metadata = FALSE) sce_filteredM3Drop10_Zhengmix8eq(metadata = FALSE)
metadata |
Logical, whether only metadata should be returned |
SingleCellExperiment
This is a scRNA-seq data set originally from Zheng et al. (2017). The data set consists of eight pre-sorted cell types (B-cells, naive cytotoxic T-cells, CD14 monocytes, regulatory T-cells, CD56 NK cells, memory T-cells, CD4 T-helper cells and naive T-cells) from Homo sapiens combined into three artificial data sets with different cell proportions. The annotated cell type (obtained by pre-sorting of the cells) is used as the true cell label. The data sets have been used to evaluate the performance of clustering algorithms in Duò et al. (2018).
For the Zhengmix4eq data set, randomly selected B-cells, CD14 monocytes,
naive cytotoxic T-cells and regulatory T-cells were combined in equal
proportions (1,000 cells per subpopulation).
The Zhengmix4uneq data set consists of four cell types, combined in
unequal proportions (1,000 B-cells, 500 naive cytotoxic T-cells, 2,000 CD14
monocytes and 3,000 regulatory T-cells). For the Zhengmix8eq data set,
all eight populations were combined in approximately equal proportions (400–600
cells per population).
For the sce_full_Zhengmix4eq, sce_full_Zhengmix4uneq and
sce_full_Zhengmix8eq data set, all genes except those with zero counts
across all cells are retained. The gene counts are unique molecular
identifiers (UMIs) counts.
The scater package was used to perform quality control of the data
(McCarthy et al. (2017)).
Features with zero counts across all cells, as well as all cells with total
count or total number of detected features more than 3 median absolute
deviations (MADs) below the median across all cells (on the log scale),
were excluded.
The sce_full_Zhengmix4eq data set consists of 3,994 cells and 15,568
features, the sce_full_Zhengmix4uneq data set of 6,498 cells and 16,443
features and the sce_full_Zhengmix8eq of 3,994 cells and 16,443 features,
respectively.
The filteredExpr, filteredHVG and filteredM3Drop10 are
further reduced data sets.
For each of the filtering method, we retained 10 percent of the original
number of genes
(with a non-zero count in at least one cell) in the original data sets.
For the filteredExpr data sets, only the genes with the highest average
expression (log-normalized count) value across all cells were retained.
Using the Seurat package, the filteredHVG data sets were filtered
on the variability of the features and only the most highly variable ones were
retained (Satija et al. (2015)). Finally, the M3Drop package was used
to model the dropout rate of the genes as a function of the mean expression
level using the Michaelis-Menten equation and select variables to retain for the
filteredM3Drop10 data sets (Andrews and Hemberg (2018)).
The scater package was used to normalize the count values, based on
normalization factors calculated by the deconvolution method from the
scran package (Lun et al. (2016)).
This data set is provided as a SingleCellExperiment object
(Lun and Risso (2017)). For further information on the
SingleCellExperiment class, see the corresponding manual.
Raw data files or the original data sets are available from
https://support.10xgenomics.com/single-cell-gene-expression/datasets.
Returns a SingleCellExperiment object.
Andrews, T.S., and Hemberg, M. (2018). Dropout-based feature selection for scRNASeq. bioRxiv doi:https://doi.org/10.1101/065094.
Duò, A., Robinson, M.D., and Soneson, C. (2018). A systematic performance evaluation of clustering methods for single-cell RNA-seq data. F1000Res. 7:1141.
Lun, A.T.L., Bach, K., and Marioni, J.C. (2016) Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 17(1): 75.
Lun, A.T.L., and Risso, D. (2017). SingleCellExperiment: S4 Classes for Single Cell Data. R package version 1.0.0.
McCarthy, D.J., Campbell, K.R., Lun, A.T.L., and Wills, Q.F. (2017): Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics 33(8): 1179-1186.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33(5): 495–502.
Zheng, G.X., Terry, J.M., Belgrader P., Ryvkin, Pl, Bent, Z.W., Wilson, R., Ziraldo, S.B., Wheeler, T.D., McDermott, G.P., Zhu, J., Gregory, M.T., Shuga, J., Montesclaros, L., Underwood, J.G., Masquelier, D.A., Nishimura, S.Y., Schnall-Levin, M., Wyatt, P.W., Hindson, C.M., Bharadwaj, R., Wong, A., Ness, K.D., Beppu, L.W., Deeg, H.J., McFarland, C., Loeb, K.R., Valente, W.J., Ericson, N.G., Stevens, E.A., Radich, J.P., Mikkelsen, T.S., Hindson, B.J., and Bielas, J.H. (2017). Massively parallel digital transcriptional profiling of single cells. Nat. Commun. 8:14049.
sce_filteredExpr10_Zhengmix4eq()sce_filteredExpr10_Zhengmix4eq()
Calculate Shannon entropy
shannon_entropy(cluster_assignments)shannon_entropy(cluster_assignments)
cluster_assignments |
A vector with cluster assignments |
The Shannon entropy of the assignment vector