WITHDRAWN: PasLINCS: Pathway Activity Signatures from LINCS L1000 Consensus Gene Signatures

• Activity Number: 182 - SPEED: New Methods in Statistical Genomics and Genetics Part 2
• Type: Contributed
• Date/Time: Monday, July 29, 2019 : 10:30 AM to 11:15 AM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Yan Ren and Siva Sivaganesan and Nicholas Clark and David Plas and Mario Medvedovic
• Companies: University of Cincinnati and University of Cincinnati and University of Cincinnati and University of Cincinnati and University of Cincinnati
• Keywords: mechanism of action; Bayesian; signed Laplacian; network kernel

Abstract

Misregulation of signaling pathway activity (SPA) is etiologic for many human diseases, and modulating activity of signaling pathways dysregulated by the disease is often the preferred therapeutic strategy. Understanding the mechanism of action of a chemical perturbagen (CP) on targeted signaling pathways is the essential first step in evaluating its therapeutic potential. The transcriptional signature of a CP provides a convenient, high information content readout of changes in the cellular state after perturbation. However, the changes of SPA are often not mediated by changes in mRNA levels of pathway constituents, rendering direct enrichment-type analyses ineffective. Using a new Bayesian approach, we identified changes in SPA with high accuracy. We use knockdown Consensus Gene Signatures in LINCS libraries to construct a novel pathway activity signature (PAS). We show that correlating PAS with a CP transcriptional signature is an effective new way of evaluating the potential of the perturbagen to modulate pathway activity. Furthermore, we show that PAS can be used to further refine network topology, optimizing the network for data mining across or within cell types.