Workshop: Inference of allele specific expression levels from RNA-Seq data

In this work, a recent method for SNV detection and genotyping from RNA-Seq data is integrated with the scalable haplotype reconstruction algorithm of and a diploid version of the Expectation Maximization (EM) algorithm for isoform expression estimation of into a pipeline for estimation of allele-specific isoform expression levels. Our pipeline does not require genome sequencing data, but can incorporate such data when available. Inferring the two haplotypes and re-mapping the reads against the diploid transcriptome resolves the above mentioned bias towards reference alleles, while the EM model improves inference accuracy by using all reads, including those that map to more than one isoform, incorporating additional sources of disambiguation information such as the distribution of RNA- Seq fragment lengths, and correcting biases introduced by library preparation and sequencing protocols.