Batch Correction with Bulk RNA-Seq

Bulk RNA-seq has become routine for unbiased transcript/gene expression quantification in samples across multiple conditions. There is currently data for literally hundreds of thousands of samples on public databases like the Gene Expression Omnibus (GEO).

We may be tempted to include data from some of these samples in our own analyses to either enrich or provide more context to the findings with our own data. However, it is well known that these data suffer from batch effects (the lab, protocol, instrument used are unfortunately the main source of variation of these data types and not the underlying biology). Under what circumstances can we include external data? What are some tools that we can employ for this purpose?

We will answer questions (like the two mentioned above) as specifically applied to the RNA-seq data related contexts of the participants during this office hour.

Series

This is part of a series of office hours (via Zoom) on particular bioinformatics topics in the coming weeks. The goal is to help researchers think through their analysis challenges and also serve as a learning opportunity. These office hours will be different from the workshops we offer in that our focus here will be on particular problems faced by participants. You will therefore have an opportunity to email details of the specific problems you are encountering after registering for a given office hour session.