A Multilayer Network Approach for Guiding Drug Repositioning in Neglected Diseases

ARIEL BERENSTEIN; MARIA PAULA MAGARIÑOS; ARIEL CHERNOMORETZ; FERNAN AGUERO

PLOS NEGLECTED TROPICAL DISEASES

PUBLIC LIBRARY SCIENCE

Lugar: San Francisco; Año: 2016 vol. 10

1935-2735

Drug development for neglected diseases has been historically hampered due to lack ofmarket incentives. The advent of public domain resources containing chemical informationfrom high throughput screenings is changing the landscape of drug discovery for these dis-eases. In this work we took advantage of data from extensively studied organisms likehuman, mouse, E. coli and yeast, among others, to develop a novel integrative networkmodel to prioritize and identify candidate drug targets in neglected pathogen proteomes,and bioactive drug-like molecules. We modeled genomic (proteins) and chemical (bioactivecompounds) data as a multilayer weighted network graph that takes advantage of bioactiv-ity data across 221 species, chemical similarities between 1.7 105 compounds and severalfunctional relations among 1.67 105 proteins. These relations comprised orthology, sharingof protein domains, and shared participation in defined biochemical pathways. We show-case the application of this network graph to the problem of prioritization of new candidatetargets, based on the information available in the graph for known compound-target associ-ations. We validated this strategy by performing a cross validation procedure for knownmouse and Trypanosoma cruzi targets and showed that our approach outperforms classicalignment-based approaches. Moreover, our model provides additional flexibility as two dif-ferent network definitions could be considered, finding in both cases qualitatively differentbut sensible candidate targets. We also showcase the application of the network to suggesttargets for orphan compounds that are active against Plasmodium falciparum in high-throughput screens. In this case our approach provided a reduced prioritization list of targetproteins for the query molecules and showed the ability to propose new testable hypothesesfor each compound. Moreover, we found that some predictions highlighted by our networkmodel were supported by independent experimental validations as found post-facto in theliterature.