Comparison and evaluation of statistical error models for scRNA-seq
Abstract Background Heterogeneity in single-cell RNA-seq (scRNA-seq) data is driven by multiple sources, including biological variation in cellular state as well as technical variation introduced during experimental processing. Deconvolving these effects is a key challenge for preprocessing workflow...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2022-01-01
|
| Series: | Genome Biology |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13059-021-02584-9 |
| _version_ | 1819261447926972416 |
|---|---|
| author | Saket Choudhary Rahul Satija |
| author_facet | Saket Choudhary Rahul Satija |
| author_sort | Saket Choudhary |
| collection | DOAJ |
| description | Abstract Background Heterogeneity in single-cell RNA-seq (scRNA-seq) data is driven by multiple sources, including biological variation in cellular state as well as technical variation introduced during experimental processing. Deconvolving these effects is a key challenge for preprocessing workflows. Recent work has demonstrated the importance and utility of count models for scRNA-seq analysis, but there is a lack of consensus on which statistical distributions and parameter settings are appropriate. Results Here, we analyze 59 scRNA-seq datasets that span a wide range of technologies, systems, and sequencing depths in order to evaluate the performance of different error models. We find that while a Poisson error model appears appropriate for sparse datasets, we observe clear evidence of overdispersion for genes with sufficient sequencing depth in all biological systems, necessitating the use of a negative binomial model. Moreover, we find that the degree of overdispersion varies widely across datasets, systems, and gene abundances, and argues for a data-driven approach for parameter estimation. Conclusions Based on these analyses, we provide a set of recommendations for modeling variation in scRNA-seq data, particularly when using generalized linear models or likelihood-based approaches for preprocessing and downstream analysis. |
| first_indexed | 2024-12-23T19:41:57Z |
| format | Article |
| id | doaj.art-609c7e20f7084346bc7c38f55ee9adcb |
| institution | Directory Open Access Journal |
| issn | 1474-760X |
| language | English |
| last_indexed | 2024-12-23T19:41:57Z |
| publishDate | 2022-01-01 |
| publisher | BMC |
| record_format | Article |
| series | Genome Biology |
| spelling | doaj.art-609c7e20f7084346bc7c38f55ee9adcb2022-12-21T17:33:38ZengBMCGenome Biology1474-760X2022-01-0123112010.1186/s13059-021-02584-9Comparison and evaluation of statistical error models for scRNA-seqSaket Choudhary0Rahul Satija1New York Genome CenterNew York Genome CenterAbstract Background Heterogeneity in single-cell RNA-seq (scRNA-seq) data is driven by multiple sources, including biological variation in cellular state as well as technical variation introduced during experimental processing. Deconvolving these effects is a key challenge for preprocessing workflows. Recent work has demonstrated the importance and utility of count models for scRNA-seq analysis, but there is a lack of consensus on which statistical distributions and parameter settings are appropriate. Results Here, we analyze 59 scRNA-seq datasets that span a wide range of technologies, systems, and sequencing depths in order to evaluate the performance of different error models. We find that while a Poisson error model appears appropriate for sparse datasets, we observe clear evidence of overdispersion for genes with sufficient sequencing depth in all biological systems, necessitating the use of a negative binomial model. Moreover, we find that the degree of overdispersion varies widely across datasets, systems, and gene abundances, and argues for a data-driven approach for parameter estimation. Conclusions Based on these analyses, we provide a set of recommendations for modeling variation in scRNA-seq data, particularly when using generalized linear models or likelihood-based approaches for preprocessing and downstream analysis.https://doi.org/10.1186/s13059-021-02584-9Single-cell RNA-seqNormalizationDimension reductionVariable genesDifferential expressionFeature selection |
| spellingShingle | Saket Choudhary Rahul Satija Comparison and evaluation of statistical error models for scRNA-seq Genome Biology Single-cell RNA-seq Normalization Dimension reduction Variable genes Differential expression Feature selection |
| title | Comparison and evaluation of statistical error models for scRNA-seq |
| title_full | Comparison and evaluation of statistical error models for scRNA-seq |
| title_fullStr | Comparison and evaluation of statistical error models for scRNA-seq |
| title_full_unstemmed | Comparison and evaluation of statistical error models for scRNA-seq |
| title_short | Comparison and evaluation of statistical error models for scRNA-seq |
| title_sort | comparison and evaluation of statistical error models for scrna seq |
| topic | Single-cell RNA-seq Normalization Dimension reduction Variable genes Differential expression Feature selection |
| url | https://doi.org/10.1186/s13059-021-02584-9 |
| work_keys_str_mv | AT saketchoudhary comparisonandevaluationofstatisticalerrormodelsforscrnaseq AT rahulsatija comparisonandevaluationofstatisticalerrormodelsforscrnaseq |