Empirical bayes analysis of sequencing-based transcriptional profiling without replicates

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• High throughput sequencing is increasingly used for transcriptome analysis.
• Sequencing offers advantages over microarrays, including reduced technical variability.
• Biological variation between replicates is often overlooked in analyses.

Purpose of the Study:

• To develop a novel method for detecting differential gene expression in high throughput sequencing data.
• To address the challenge of biological variation, particularly when replicates are limited.

Main Methods:

• An empirical Bayes method, Analysis of Sequence Counts (ASC), was developed.
• ASC borrows information across sequences to model sample variation.
• It establishes a prior distribution for sample variation.

Main Results:

• ASC accounts for biological variation even with few or no replicates.
• The method is less biased towards highly expressed sequences compared to proportion tests.
• ASC identifies more genes with greater log fold change at lower abundance.

Conclusions:

• ASC integrates biological and statistical significance for differential expression.
• It estimates the posterior mean of log fold change and false discovery rates.
• An R implementation is available for public use.