Batch effects and the effective design of single-cell gene expression studies
Po-Yuan Tung1, John D Blischak1,2, Chiaowen Joyce Hsiao1
1Department of Human Genetics, University of Chicago, Chicago, Illinois, USA.
Scientific Reports
|January 4, 2017
Summary
Technical variation in single-cell RNA sequencing (scRNA-seq) using Fluidigm C1 is substantial. Unique molecular identifiers (UMIs) do not fully correct for amplification bias, impacting gene expression accuracy.
Area of Science:
- Genomics and transcriptomics
- Molecular biology
- Biotechnology
Background:
- Single-cell RNA sequencing (scRNA-seq) offers high-resolution gene expression analysis.
- Understanding experimental noise sources in scRNA-seq is crucial for accurate interpretation.
- Technical variation during sample processing can confound biological insights.
Purpose of the Study:
- To investigate technical variation in scRNA-seq data generated by the Fluidigm C1 platform.
- To assess the impact of unique molecular identifiers (UMIs) on correcting amplification bias.
- To propose a framework for optimizing scRNA-seq experimental design.
Main Methods:
- Utilized Fluidigm C1 platform for single-cell capture and processing.
- Incorporated unique molecular identifiers (UMIs) to all samples.
- Analyzed gene expression data from three human induced pluripotent stem cell (iPSC) lines across technical replicates.
Main Results:
- Genotype was the primary driver of variation, but significant technical variation existed between replicates.
- UMI-based molecule counts were influenced by both biological and technical factors.
- UMI counts are not a fully unbiased estimator of absolute gene expression levels.
Conclusions:
- Technical variation is a significant factor in scRNA-seq data quality, even with UMIs.
- The Fluidigm C1 platform introduces notable technical noise.
- A robust framework is needed to account for technical variation in scRNA-seq studies.
