Alternative splicing of RNA allows a single gene to create diverse proteins. Misregulation of splicing is a characteristic of many diseases, such as cancer. Accurate quantification of splicing is important for allowing researchers to understand the state of a given sample. The projects described here concern comprehensive methods for quantifying alternative splicing, to provide insight into global splicing patterns and individual splicing events.

MESA is a tool we have developed for this quantification, giving tables of percent-spliced (PS) values for all splicing events in a dataset. It detects more events, with a better sensitivity for detecting alternative splicing differences, than existing tools. The PS tables can be used to generate splicing signatures, such as the signature of a disease. The signature for a particularly difficult to treat subtype of breast cancer, known as triple-negative breast cancer (TNBC) was generated from a large dataset of patient tumors. The signature was found to be prognostic, as high similarity to the signature was associated with poor survival.
I plan to generate a set of PS value tables, and splicing signatures, from several datasets of cancer samples and healthy tissue samples. I will provide these for researchers to compare to their own samples. Along with this, I will develop a set of summary statistics to describe global splicing patterns within a sample. I will also continue to develop a method for finding the statistical significance of how well a sample matches a particular signature.

Finally, I will develop methods for quantifying splicing and finding signatures from long-read RNA sequencing data. This first involves comparing PS values between short and long-read data. Then, I will develop a tool called DICE, that will extend MESA’s algorithm to look at pairs of splicing events, in order to capture more information about the full-length isoform. I will apply this method to single-cell and neural tissue data, in order to examine alternative splicing differences that are difficult to detect with current methods.

Event Host: Dennis Mulligan, Ph.D. Student, Biomolecular Engineering & Bioinformatics 

Advisor: Angela Brooks

Join us in person or on Zoom: https://ucsc.zoom.us/j/3532726499?pwd=bG9zbG1mY2lvWHB5RW9BSUVwZyt0UT09

Passcode: RNA