Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression

Нормализация и стабилизация дисперсии данных одноклеточного RNA-секвенирования с использованием регуляризованной отрицательно-биномиальной регрессии
Rahul Satija, Christoph Hafemeister, Rahul Satija
2019-12-01

Pearson residualsregularized negative binomial regressionsctransformsingle-cell RNA-seqvariance stabilization
Single-cell RNA-seq (scRNA-seq) data exhibits significant cell-to-cell variation due to technical factors, including the number of molecules detected in each cell, which can confound biological heterogeneity with technical effects. To address this, we present a modeling framework for the normalization and variance stabilization of molecular count data from scRNA-seq experiments. We propose that the Pearson residuals from "regularized negative binomial regression," where cellular sequencing depth is utilized as a covariate in a generalized linear model, successfully remove the influence of technical characteristics from downstream analyses while preserving biological heterogeneity. Importantly, we show that an unconstrained negative binomial model may overfit scRNA-seq data, and overcome this by pooling information across genes with similar abundances to obtain stable parameter estimates. Our procedure omits the need for heuristic steps including pseudocount addition or log-transformation and improves common downstream analytical tasks such as variable gene selection, dimensional reduction, and differential expression. Our approach can be applied to any UMI-based scRNA-seq dataset and is freely available as part of the R package sctransform, with a direct interface to our single-cell toolkit Seurat.
1
Pearson residuals from the model reduce technical effects while preserving biologically meaningful cell-to-cell heterogeneity.
2
Pooling information across genes with similar abundances prevents overfitting and produces more stable parameter estimates than unconstrained negative binomial modeling.
3
Regularized negative binomial regression uses cellular sequencing depth as a covariate to normalize UMI-based scRNA-seq counts and stabilize their variance.
4
The method avoids heuristic pseudocount addition and log-transformation, while improving variable gene selection, dimensionality reduction, and differential expression analyses.
5
The procedure is applicable to any UMI-based scRNA-seq dataset and is implemented in the freely available sctransform R package with Seurat integration.

UMI-based single-cell RNA-seq molecular count data

Normalization and variance stabilization of scRNA-seq counts by removing sequencing-depth-related technical variation while preserving biological heterogeneity

Publication Details
Publication Date
2019-12-01
Journal
Publisher
ISSN
Cited by
4884
Access Type
Author Information
Authors
Rahul Satija
Christoph Hafemeister
Rahul Satija
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%