Research group coordinated by Prof. Giulio Vistoli
The scientific activity of Drug Design Lab (DDL) has been mainly focused on computational approaches as applied to medicinal chemistry in its broadest sense ranging from homology modeling and virtual screening to ADME predictions or physicochemical profiling. The production involved several thematic areas, which have been undertaken utilizing appropriate known approaches or developing novel computational methodologies with a view to successfully pursuing the planned objectives.
Taken together, the DDL’s scientific activity allows three main research interests to be identified:
1) Modeling of unresolved proteins and analysis of their interaction capacities through docking simulations and/or molecular dynamics.
This research line involves both transmembrane and soluble proteins whose three-dimensional structure is generated by homology techniques. With regards to transmembrane proteins, great effort has been focused on the modeling of medicinally relevant GRPC models. In particular, DDL group has developed a fragmental approach to account for the local peculiarities of GPCR structures thus avoiding the generation of models too similar to the reference template. In detail, the group has developed a computational method to simulate the dynamic behavior of GPCRs by suitably exploiting the conformational variability of the proline containing transmembrane helices. This research line also involves ion channels with particular interest to neuronal nicotinic receptors and thermo transient potential receptors (TRPs). Concerning the soluble proteins, modeling studies involved both metabolically relevant enzymes (such as carboxyleaterases or serum carnosinase) and medicinal targets (such as histone deacetylases or seladin-1). Very often, these modeling studies were paralleled by targeted MD simulations with a view to revealing the different fate of substrates and products or, as in the case of serum carnosinase, to investigating the allosteric effect of the citrate ion. It is worth noting that several homology modeling studies were performed to support the rational design of improved carnosine analogues as potent, selective and bioavailable carbonyl quenchers, a project on which the group is heavily involved in the last few years.
2) Physicochemical profiling through the purposely developed property space approach.
This research line includes the methodological development of the concept of property space which is based on the variability of conformer-dependent physicochemical properties. The approach allows a better characterization of the dynamic behavior of a given flexible molecule to be better characterized by accounting not only for its conformational mobility but also for its property variability. The property space concept was developed by investigating the property profiles of a single well-known and biologically relevant molecule, acetylcholine, and then found fertile applications to account for the entropic component in the ligand recognition processes. The property space concept also lead to the definition of novel molecular descriptors (such as property range and molecular sensitivity) which found encouraging applications in QSAR analyses and, in particular, in ADME predictions.
3) Software development.
The DDL group is involved in VEGA ZZ development since 1996 and the first public command line version (V. 1.1) was released in 1998. VEGA ZZ is a complete suite of programs that includes many features to support molecular modeling studies. Starting from 1.3 release (2001), the program has a graphic interface with 3D OpenGL output. Starting from 2.0 release (2004), the program is distributed through a fully-automated license manager that allows the VEGA ZZ users to be monitored. During these years, the program has been constantly enriched by new features, scripts and plug-ins. Now, VEGA ZZ is a worldwide distributed suite of programs as clearly demonstrated by 12778 registered users and 18151 active licenses.
The group has also developed other pieces of software for exhaustive blind docking analyses (BioDock), for protein-protein and protein-DNA docking studies (Escher NG) and for the virtual screening (GriDock).