The combined 2013 and 2014 Community Structure-Activity Resource (CSAR) Competition (www.csardock.org) has concluded, and the results are in.
The CSAR competition is an annual competition organized by Drs. Heather Carlson and James Dunbar at the University of Michigan. This year, a total of 14 international teams competed. Researchers at the University of Pittsburgh submitted predictions for the five phases of competition and have come out on top across the board.
Professor Carlos J. Camacho and graduate student Matthew Baumgartner, of University of Pittsburgh’s Department of Computational and Systems Biology, used a number of in-house computational methods to solve the five challenges. The nature of the challenges range from: predicting which proteins bind to a given compound when given only the primary sequence of amino acids, choosing the correct binding pose of a compound out of a set of 200 bad poses, and ranking a set of compounds based binding affinity. The Camacho Lab submissions in all categories were among the best predictions. Due to their performance, the Lab was invited to submit a paper detailing their efforts to the Journal of Chemical Information and Modeling. The expected publication date is late 2015.