Splice Site Algorithms for Clinical Genomics

Abstract

To fully interpret variants in the context of clinical genomics, as outlined by the ACMG interpretation guidelines, variants near canonical splice boundaries must be evaluated for their potential to disrupt gene splicing and thus be classified as a gene-damaging mutation. Five splicing methods have been canonized for this purpose in the clinical testing market: GeneSplicer, MaxEntScan, NNSplice, and Position Weight Matrix (PWM). Although these algorithms vary wildly in their performance characteristics, such as sensitivity and specificity, they are treated as black-box oracles on equal footing when being used by variant curators to classify variants.

In this presentation, we will review these algorithms and their technical strengths and weakness from the perspective of clinical variant interpretation. Additionally, we will present our approach to splice site prediction within VarSeq, demonstrating how we improve on the status quo by employing these algorithms both in the batch annotation and genomic visualization contexts. Finally, we will demonstrate how these methods can be leveraged during the interpretation of variants following the ACMG guidelines within our upcoming product, VSClinical.

About the Presenter

Nathan Fortier

Nathan Fortier joined the Golden Helix development team in June of 2014 and is a Senior Software Engineer and Field Application Scientist. Nathan obtained his Bachelor’s degree in Software Engineering from Montana Tech University in May 2011, received a Master’s degree in Computer Science from Montana State University in May 2014, and received his Ph.D. in Computer Science from Montana State University in May 2015. Nathan works on data curation, script development, and product code. When not working, Nathan enjoys hiking and playing music.