The importance of taking into account protein f
lexibi
lity in drug design and virtua
l ligand screening (VS) has been wide
ly debated in the
literature, and mo
lecu
lar dynamics (MD) has been recognized as one of the most powerfu
l too
ls for investigating intrinsic protein dynamics. Neverthe
less, deciphering the amount of information hidden in MD simu
lations and recognizing a significant minima
l set of states to be used in virtua
l screening experiments can be quite comp
licated. Here we present an integrated MD鈥揊LAP (mo
lecu
lar dynamics鈥揻ingerprints for
ligand and proteins) approach, comprising a pipe
line of mo
lecu
lar dynamics, c
lustering and
linear discriminant ana
lysis, for enhancing accuracy and efficacy in VS campaigns. We first extracted a
limited number of representative structures from tens of nanoseconds of MD trajectories by means of the k-medoids c
lustering a
lgorithm as imp
lemented in the BiKi Life Science Suite (
lass="extLink">http://www.bikitech.com [accessed July 21, 2015]). Then, instead of applying arbitrary selection criteria, that is, RMSD, pharmacophore properties, or enrichment performances, we allowed the linear discriminant analysis algorithm implemented in FLAP (ldiscovery.com" class="extLink">http://www.moldiscovery.com [accessed July 21, 2015]) to automatically choose the best performing conformational states among medoids and X-ray structures. Retrospective virtual screenings confirmed that ensemble receptor protocols outperform single rigid receptor approaches, proved that computationally generated conformations comprise the same quantity/quality of information included in X-ray structures, and pointed to the MD鈥揊LAP approach as a valuable tool for improving VS performances.