A quantitative profiling tool for diverse genomic data types reveals potential associations between chromatin and pre-mRNA splicing

ISAAC KREMSKY; NICOLAS BELLORA; EDUARDO EYRAS

Barcelona

Jornada; GRIB EXPO - The Big Data Challenge; 2014

Research Program on BIomedical Informatics - Universitat Pompeu Fabra

A quantitative profiling tool for diverse genomic data typesreveals potential associations between chromatin andpre-mRNA splicingIsaac Kremsky, Nicolas Bellora, and Eduardo EyrasWe have developed bioinformatics software to generateprofiles of occurrences of genomic data around a set ofreference positions. Mulitiple profiles can be generated ona single plot; for pairwise profiling, the software calculatesposition-specific P-values using Fisher's exact test todetermine if differences between the two profiles arestatistically significant. P-values can also be calculated tocompare regions of a single profile. Quantitative,controlled studies of one or more genome-wide datasetscan therefore be performed with this software. Potentialbiases can be systematically checked and in some caseseliminated entirely. Applicable data types includemappability. motifs, CLIP-seq, ChIP-seq, iCLIP-seq, andGro-seq Insights and extensions of published findingsbased on several of these data types have been gleaned.In particular, a potential link between the splicing factorSRSF1 and the DNA-binding protein Ctcf, already knownto play a role in alternative splicing, has been identified.