At present, we are focusing our efforts in three main directions:
- Developing new algorithms to accelerate the simulation of processes of pharmaceutical and biological relevance, such as the interaction of proteins and drug candidates;
- Improving the description of systems from the nanoscale to the mesoscale by means of continuum electrostatics and advanced implicit solvent methods, with particular attention devoted to the description and calculation of molecular surfaces;
- Developing highly scalable machine learning methods for the analysis of large data produced by microsecond long molecular dynamics simulation.
The group is at present composed by three post-docs and one PhD students, in addition to the group leader. We make our implementations on multicore CPU workstations and an internal cluster and on mixed CPU-GPU architectures. We have also access to a Blue Gene machine in the largest Italian public supercomputing centre (CINECA) thanks also to the PRACE initiative.
- In 2010: 5-year NIH grant (as Co-Investigator) “DelPhi: Software for Electrostatic Modeling of Biomolecules and Objects”.
- In 2013: MOST-MAE bilateral Italy-Israel grant (as PI) “A comprehensive structure-based approach for identifying novel pathogen-specific drug targets"
- Honig’s Lab (Columbia University, USA) - continuum electrostatics modelling, development and improvement of the Poisson-Boltzmann solver DelPhi;
- CNR-IMATI (ITA) - geometric aspects of molecular surfaces and data analysis;
- Kosloff Lab (ISRAEL) - a bilateral Italy-Israel grant is funding a shared project, aimed at finding new strategies against resistant pathogens.
Walter Rocchia is co-founder of BiKi Technologies, a spin-off of IIT incorporated in May 2014. BiKi comprises a multidisciplinary team of software engineers, computational chemists and biochemists working at the intersection of basic and applied research, providing pharmaceutical and biotech companies with innovative computational solutions. (http://www.bikitech.com/)