Faculty Advisor or Committee Member

Amity Manning, Reader

Faculty Advisor or Committee Member

Dmitry Korkin, Advisor




The identification of alternative splicing in the human genome elucidated the potential to several enduring genomic questions. Not only could this phenomenon explain why organism complexity was not at all correlated with the genome size, or explain how an organisms could be affected by experience and environment at the molecular level, but it was perhaps the most flexible and adaptive regulatory mechanism identified to date. While the pathogenic aberrations of this mechanism have generally been readily investigated and identified as potential therapeutic targets, its meditative or advantageous instances have largely not been considered. Initiated exon skipping has been shown to have therapeutic effects in Muscular Dystrophy animal models and even in vitro human muscle cells (Aartsma-Rus, Annemieke, et al, Human Molecular Genetics 2003, McClorey, G., et al, Neuromuscular disorders, 2006). However, the consideration that this process may be occurring endogenously in human cells and contributing to other complex diseases has remained largely ignored. In this work, we have undertaken the first large-scale statistical examination of alternatively spliced variants between the tissues of diseased and normal patients. We hypothesize that there are endogenous alternative splicing events occurring in these tissues that purposefully mediate mutative damage and contribute to the differentiation between diseased and healthy phenotypes. By integrating data from several different sources and employing statistic and machine learning models, we have identified significant differences in gene characteristics between canonical and spliced variants correlated with changes in clinical outcomes. We conclude that this evidence supports our hypothesis that alternative splicing can be positively driven to mediate genetic damage. Expression of these genetically damaged and canonically spliced variants is clearly implicated in diseased tissue and poor clinical outcomes.


Worcester Polytechnic Institute

Degree Name



Bioinformatics and Computational Biology

Project Type


Date Accepted





Alternative Splicing, Cancer, Machine Learning, mRNAseq, Mutation, Statistics