SPILLO-PBSS software

Prediction of off-target-based toxicologic effects with the innovative software SPILLO-PBSS

Alessandro Di Domizio
Founder of SPILLO Project

The risk of low selectivity usually represents a serious problem when designing new bioactive molecules; particularly, because similar binding sites could appear even within proteins sharing low sequence and/or structural homology.

On these grounds, the possibility to predict off-target interactions of a therapeutic agent, starting from a well-defined reference binding site, could be of significant support in the Drug Discovery and Development process.

The innovative software SPILLO-Potential Binding Sites Searcher (SPILLO-PBSS 1 ) will be presented. Unlike the case of traditional structure-based approaches, SPILLO-PBSS is able to detect potential binding sites within protein 3D-structures, even when these are highly distorted compared to a suitable binding conformation. In this way the risk of missing potentially relevant off-target proteins is greatly reduced. Morover, SPILLO-PBSS is fast enough for use even on a proteome-wide scale, with relatively limited computational resources, thus representing a valuable support at many stages of Drug R&D (prediction and/or clarification of side effects, polypharmacology, drug repositioning).

The participants will be given a practical opportunity to try SPILLO-PBSS on individual proteins, working out some representative examples, while getting familiar with the software. However, particular attention will be given to the importance of preliminar feasibility studies as well as the analysis of the results obtained.

The participants will then be able to apply SPILLO-PBSS to any potential antimalarial biomolecule identified during the previous days of the workshop, in order to predict their possible side-effects on humans. Potential off-target proteins of these compounds will be detected by virtual screenings of a subset of human protein kinases.

Lastly, a further opportunity will be offered to launch virtual screenings of the whole human structural proteome available (11,500+ protein 3D-structures, without sequence redundancies, from the Protein Data Bank (PDB)) and to retrieve the results at a later date.

1) A. Di Domizio et al (2014). SPILLO-PBSS: Detecting hidden binding sites within protein 3D-structures through a flexible structure-based approach. J. Comp. Chem. 35 (27): 2005-2017.