Development of an in silico transcriptional drug repositioning methodology for immune cell modulation

Abstract

Cancer research has made great advances during the last decades in the understanding of the mechanisms that govern the course of the disease. One key factor for this advance has been the comprehension of the role of the immune system and its interplay with tumoral cells. Thanks to this, the view of the disease has shifted from a cancerous cell-centric view, towards a more holistic view in which other cells crosstalk with the tumor and determine its course. This paradigm has allowed the discovery, development and approval of compounds with immunomodulatory capabilities. Computational biology has been extensively employed to analyze large-scale datasets to decipher the interplay mechanisms employed by the tumor and its microenvironment. This project aims to explore potential strategies to exploit immune modulation in the context of cancer and other immune-driven diseases. The principal aim of this thesis is the design of a computational framework to allow hypothesis generation of compounds capable of modulating immune cell activity from gene expression signatures. To this end, we have compiled, annotated and curated a large collection of immune gene expression signatures derived from the literature or generated in-house and a collection of drug-induced gene expression profiles from the CMap L1000 consortium. By employing computational methodologies based on GSEA, the signatures and the drug-induced profiles are compared and scored according to their degree of similarity. These scores are then employed to generate an immune signature-drug association database available to query for compounds with potential immunomodulating capabilities. To facilitate the access to the database we have developed DREIMT (Drug REpositioning for IMmune Transcriptome; www.dreimt.org), a web tool from which queries to the database can be performed. Furthermore, DREIMT has a built-in tool to generate hypotheses of immunomodulatory compounds from user-provided immune signatures and a tool to compare a user-provided immune signature to those employed in the database