Computational drug design

My PhD thesis and current work offers some methodological solutions to accelerate in silico drug design process. The issues dealt with include time efficient, accurate partial atomic charge assignment scheme of small organic (candidate drug) molecules, rapid screening of large databases of small molecules against any target protein and prediction of sites of metabolism of xenobiotics against any isoforms of Cytochrome P-450 enzyme etc.

A brief summary of my software development activities can be found here.

Apart from method development in drug design, I had also been actively involved in collaborative research on the design of lead-like molecules on gelsolin and SARS CoV2 targets (Fig. 1). SARS CoV2 small molecules design project was carried out by three research groups, who independently conducted different parts of the pipeline (Fig. 1), which included rapid virtual screening using several methods (RASPD+), ligand similarity searches and ROCS ligand pharmacophore screens), molecular docking with three independent methods (GOLD, Glide, FRED), molecular dynamics simulation of docked complexes (with analysis with our MD-IFP workflow), and ADMET screening (Qikprop). The results were then compared to identify consensus compounds that could be taken through the further steps of the pipeline (e.g. MD simulation and ADMET assessment). Currently, the experimental evaluation of our designed lead-like molecules is under process.

Figure 1: The computational pipeline employed in designing small molecule inhibitors against SARS CoV2 targets