Many mutations that cause cancer and other diseases may not be detected since they look like silent mutations that do not change the encoded proteins. One fundamental mechanism that apparently silent mutations can severely disrupt is alternative splicing. We introduce a tool that scores mutations based on models of proteomes generated using aberrant splicing predictions.  The tool leverages a highly accurate neural network that predicts splice sites within arbitrary mRNA sequences, a greedy transcript constructor that considers alternate arrangements of splicing blueprints, and an algorithm that grades the functional divergence between proteins based on evolutionary conservation. we show strong enrichment of predicted deleterious mutations across pan-cancer drivers. We also achieve improved patient survival estimation using a proposed set of novel cancer-involved genes. Ultimately, this pipeline enables accelerated insight-gathering of sequence-specific consequences for a class of understudied mutations and provides an efficient way of filtering through massive variant datasets – functionalities with immediate experimental and clinical applications in any disease.  

UNMET NEED
Understanding mutations that cause diseases but look silent.
 
OUR SOLUTION
A new tool that can detect which apparently silent mutation effect splicing and thus has significant effect on the produced protein. 

APPLICATIONS
Our pipeline should be very helpful for any type of objective (diagnoss, pateints classification, development of new therapies, screen) in cancer and any other type of disease

STATUS
A tool has been developed and is ready to be used. 

INTELLECTUAL PROPERTY
A patent application was submitted.
 
REFERENCES
Nicolas Lynn, Tamir Tuller. Detecting and understanding meaningful cancerous mutations based on computational models of mRNA splicing. Under review