1. Systematic Identification of Pharmacological Targets from Small-Molecule Phenotypic Screens
- Author
-
Chung Hwee Thiam, Hoeke A. Baarsma, Monica Campillos, Corinna Montrone, Xueping Liu, Melanie Königshoff, Julia-Sophie Heier, Barbara Brauner, Veronique Angeli, Sven Duscha, Andreas Ruepp, and Gisela Fobo
- Subjects
0301 basic medicine ,Clinical Biochemistry ,Druggability ,Computational biology ,Biology ,Bioinformatics ,Biochemistry ,Cell Line ,Small Molecule Libraries ,Mice ,03 medical and health sciences ,Drug Discovery ,Animals ,Humans ,Drug Target-phenotype Relations ,High-throughput Chemical Screens ,Target De-convolution ,Target Prediction ,Molecular Targeted Therapy ,Wnt Signaling Pathway ,Molecular Biology ,Pharmacology ,Wnt signaling pathway ,Small molecule ,Phenotype ,High-Throughput Screening Assays ,Wnt Proteins ,030104 developmental biology ,A549 Cells ,Molecular Medicine ,Classical pharmacology ,Identification (biology) - Abstract
Phenotypic drug discovery offers some advantages over target-based methods, mainly because it allows drug leads to be tested in systems that more closely model distinct disease states. However, a potential disadvantage is the difficulty of linking the observed phenotype to a specific cellular target. To address this problem, we developed DePick, a computational target de-convolution tool to determine targets specifically linked to small-molecule phenotypic screens. We applied DePick to eight publicly available screens and predicted 59 drug target-phenotype associations. In addition to literature-based evidence for our predictions, we provide experimental support for seven predicted associations. Interestingly, our analysis led to the discovery of a previously unrecognized connection between the Wnt signaling pathway and an aromatase, CYP19A1. These results demonstrate that the DePick approach can not only accelerate target de-convolution but also aid in discovery of new functionally relevant biological relationships.
- Published
- 2016
- Full Text
- View/download PDF