Research

Development of computational tools to identify and quantify alternative splicing in short- and long-read RNA-seq data.

We developed a tool called Junction Based Analysis of Alternative Splicing Events (JuncBASE) which was originally described and used in identifying regulated alternative splicing events in Drosophila (Brooks et al., Genome Research 2011). JuncBASE is a computational tool that can analyze alternative splicing from RNA-seq data and integrate unannotated/novel splicing detection, classification, and quantification. We are developing another tool, MESA, for the analysis of splicing in short-read RNA-seq that allows for fast, scalable, and easy-to-interpret analysis of splicing.

We also developed a computational tool called FLAIR (Full-Length Alternative Isoform Analysis of RNA) to produce confident transcript isoforms from long-read RNA-seq data with the aim of alternative isoform detection and quantification. With an increase in the usage of long-read RNA-seq, there is a growing need for a systematic evaluation of this approach. We are part of an international community effort called the Long-read RNA-seq Genome Annotation Assessment Project (LRGASP) to perform such an evaluation. We are extending FLAIR to incorporate sequence variation, RNA editing, and RNA modification in isoform detection as well as detection of complex gene fusions from long-read sequencing data.

RNA alterations in cancer and other diseases

Using computational tools we and others develop, we aim to fully characterize the output of cancer genomes, the transcriptome, to better understand the genetic basis of cancer. We are particularly interested in RNA splicing alterations in cancer, particularly those caused by splicing factor mutations such as U2AF1 and SF3B1. We are also generally interested in alterations in RNA processing in disease.

High-throughput phenotyping of gene variants

Recent cancer genome sequencing efforts have identified millions of somatic mutations, posing a huge challenge in determining which mutations are functioning in tumorigenesis versus “passenger” mutations with no impact on gene function. This problem exists even in clinically actionable genes, such as EGFR, where it is unclear if rare somatic variants impact gene function. To address this challenge, we developed a high-throughput approach to experimentally test the functional impact of gene variants called expression-based variant impact phenotyping (eVIP), where we were able to distinguish impactful mutations from neutral, likely “passenger” mutations without any prior knowledge of gene function. We extended this work to eVIP2 and applied it to identify specific pathways affected by gene variants. We are working to extend this work to other forms of gene variants such as spliced isoforms.

Videos

Brooks Lab Introduction (2019)

Pan-cancer Analysis of Whole Genomes Project

eVIP: Expression-based variant impact phenotyping

Skip to toolbar