Speaker: Debosmita Kundu

Title: A Branching Process Model to Account for PCR Duplicated *-Seq Data

Abstract: Digital quantification by sequencing-based genetic assays is ubiquitous. One problem is the need to amplify sequences by Polymerase Chain Reaction (PCR). PCR increases variance and distorts abundances, particularly for low input. We develop a duplicate-aware branching process model for the fragment sampling, amplification, and sequencing steps underlying digital quantification. Model parameters have been estimated using Maximum Likelihood Estimation and Generalized Method of Moments. We will identify experimental conditions where the BP model is a consequentially better fit to count data from *-seq assays than competing methods.