Latent Dirichlet Model to Compare Expressed Isoform Proportions to a Reference Panel

• Activity Number: 415 - Statistical Methods for Gene Expression and RNA-Seq Analysis
• Type: Contributed
• Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Sean McCabe* and Andrew B Nobel and Michael Love
• Companies: University of North Carolina at Chapel Hill and University of North Carolina at Chapel Hill and UNC-Chapel Hill
• Keywords: Isoform Expression; Variational Bayes; Latent Variable

Abstract

The proportion of RNA isoforms (splice variants) expressed for a given gene has been associated with disease states in cancer, retinal diseases, and neurological disorders. Examination of isoform proportions can help determine biological mechanisms, however these often require a per-gene investigation of splicing patterns. Leveraging large public datasets produced by genomic consortia as a reference, we can compare splicing patterns in a dataset of interest with those of a reference panel in which samples are divided into distinct groups (tissue of origin, disease status, etc). We employ a latent Dirichlet model with Dirichlet Multinomial observations to compare expressed isoform proportions in datasets to an independent reference panel. We use a variational Bayes procedure to estimate posterior distributions for the reference panel’s sample group membership and identify sets of genes that relate to the reference panel similarly. Using the Genotype-Tissue Expression (GTEx) project as a reference dataset, we evaluate our model on simulated and real RNA-seq datasets to determine tissue type classifications of genes from an independent study.