Gene length and detection bias in single cell RNA sequencing protocols

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for transcriptomic profiling, enabling cell type discovery and tissue development insights.
• Technical challenges in scRNA-seq include amplification biases due to limited starting material.
• Unique molecular identifiers (UMIs) mitigate amplification biases by tagging individual molecules for accurate transcript abundance estimation.

Purpose of the Study:

• To investigate the impact of gene length bias in scRNA-seq across diverse datasets.
• To compare gene length bias between full-length and UMI-based scRNA-seq protocols.
• To assess the potential for combining full-length and UMI data for biological discovery.

Main Methods:

• Analysis of multiple scRNA-seq datasets with varying capture technologies, library preparations, cell types, and species.
• Comparison of gene detection rates and dropout events between full-length and UMI-based protocols.
• Examination of gene length distribution in datasets detected exclusively by either full-length or UMI methods.

Main Results:

• Full-length scRNA-seq protocols exhibit gene length bias, similar to bulk RNA-seq, with shorter genes showing lower counts and higher dropout.
• UMI-based scRNA-seq protocols do not show gene length bias, maintaining uniform dropout rates across gene lengths.
• Genes detected only in UMI datasets were shorter, while those detected only in full-length datasets were longer, using mouse embryonic stem cell data.

Conclusions:

• The choice of scRNA-seq protocol significantly influences gene detection rates and introduces gene length bias in full-length datasets.
• UMI-based protocols effectively eliminate gene length bias, offering more uniform gene detection.
• Combining full-length and UMI data can overcome individual protocol limitations and enhance the understanding of cellular expression patterns.