TDR Targets: integrated chemogenomic mining of pathogen genomes for drug discovery

LANDABURU, LIONEL URÁN; BERENSTEIN AJ; BERRIMAN M; ARIEL CHERNOMORETZ; BIVORT HAIEK F; DHANASEKARAN SHANMUGAM; VIDELA S; ZAROWIECKI M; FERNAN AGÜERO

Berlin

Simposio; 21st-Century Drug Discovery and Development for Global Health (S3); 2018

Keystone Symposia

The volume of biological, chemical and functional data deposited in the public domain is growing rapidly, thanks to next generation sequencing, and highly-automated screening technologies. However, there is still a large data imbalance between model, well-funded organisms and pathogens causing neglected diseases (NDs). We developed a chemogenomics resource, (TDR Targets, tdrtargets.org), that aims to organize and integrate heterogeneous large datasets with a focus on drug discovery for human pathogens. The database also hosts chemical and genomic data from other organisms to leverage data for comparative and inference-based queries. One of the major impacts of TDR Targets is to facilitate target and chemical prioritizations by allowing users to formulate complex queries across diverse query spaces.In this communication we will highlight new data and functionality updates in TDR Targets. In this release, the database has been updated to integrate data on >2 million bioactive compounds; 20 pathogen genomes; and 30 complete genomes from model organisms and other related pathogens. Furthermore, the data was also used to populate a recently developed network model (Berenstein AJ, 2016) to produce i) a novel druggability metric for targets based on the connectivity in the network to bioactive compounds, ii) to guide new prioritization strategies for both targets and compounds, and iii) to visually aid in the navigation across target/compound spaces in the web interface. This network model connects protein (target) nodes to compounds, based on curated bioactivity annotations. It also connects proteins to other proteins based on shared annotations, and compounds to other compounds based on chemical similarity and substructure metrics. This chemogenomic network facilitates a number of inferences, such as inferring plausible targets for orphan drugs or candidate compounds for orphan targets. Funded by ANPCyT-Glaxo Argentina (PICTO-Glaxo-2013-0067).