Inference of Gene Coexpression Networks from Bulk-Based RNA-Sequencing Data.

Gene coexpression networks (GCNs) are useful tools for inferring gene functions and understanding biological processes when properly constructed. Traditional microarray analysis is being more frequently replaced by bulk-based RNA-sequencing as a method for quantifying gene expression. This new technology requires improved statistical methods for generating GCNs.

Rab11b-mediated integrin recycling promotes brain metastatic adaptation and outgrowth.

Breast cancer brain metastases (BCBM) have a 5-20 year latency and account for 30% of mortality; however, mechanisms governing adaptation to the brain microenvironment remain poorly defined. We combine time-course RNA-sequencing of BCBM development with a Drosophila melanogaster genetic screen, and identify Rab11b as a functional mediator of metastatic adaptation. Proteomic analysis reveals that Rab11b controls the cell surface proteome, recycling proteins required for successful interaction with the microenvironment, including integrin β1.

Inference of Gene Co-expression Networks from Single-Cell RNA-Sequencing Data.

Single-cell RNA-Sequencing is a pioneering extension of bulk-based RNA-Sequencing technology. The "guilt-by-association" heuristic has led to the use of gene co-expression networks to identify genes that are believed to be associated with a common cellular function. Many methods that were developed for bulk-based RNA-Sequencing data can continue to be applied to single-cell data, and several of the most widely used methods are explored.

DiPhiSeq: robust comparison of expression levels on RNA-Seq data with large sample sizes.

Motivation: In the analysis of RNA-Seq data, detecting differentially expressed (DE) genes has been a hot research area in recent years and many methods have been proposed. DE genes show different average expression levels in different sample groups, and thus can be important biological markers. While generally very successful, these methods need to be further tailored and improved for cancerous data, which often features quite diverse expression in the samples from the cancer group, and this diversity is much larger than that in the control group.