Saeed Izadi

Scientist, Genentech, San Francisco, CA (2018-now)
Postdoc, Genentech, San Francisco, CA (2016-2018)
Research Assistant, Department of Computer Science, Virginia Tech, Blacksburg, VA (2012-2016)


Ph.D., Engineering Science and Mechanics, Virginia Tech, Blacksburg, VA, 2016
M.Sc., Aerospace Engineering, Amirkabir University of Technogoly, Tehran, Iran, 2011
B.Sc., Mechanical Engineering, Iran University of Science and Technology, Tehran, Iran, 2008


Dr. Alexey V. Onufriev, Department of Computer Science and Physics


Different approaches to parametrize explicit water models (OPC and OPC3)
Accurate yet computationally efficient description of the solvent environment is essential for realistic biomolecular modeling. In this project we developed a general framework for constructing water models based on the Optimal Point Charge Approximation concept we previously introduced. The new parametrization technique was used to build two new explicit water models: 4-point OPC and 3-point OPC3.

OPC parameter files:
OPC and OPC3 can be easily incorporated into any package the supports TIP4P and TIP3P, the corresponding parameters (including bond length and angles) need to be modified accordingly to match that of OPC models. OPC and OPC3 force field paramater files have been made available in 2015 and higher versions of the Amber package. The GROMACS parameter files are available here. Also, note that long-range correction to LJ should be applied when using OPC. More instructions are provided here.

Multi-scale approaches to speed up the computation of electrostatic interactions in large-scale simulations (GB-HCPO)
Appropriate treatment of electrostatic interactions are of fundamental importance in MD simulations of biological systems. Due to computational costs, however, calculation of these pairwise electrostatic interactions can become intractable in realistic biomolecular systems containing large numbers of atoms. The hierarchical charge partitioning (HCP) approximation computes the long-range component of intermolecular interactions in an approximate manner to speed up the computations. My task in this project is to develop computational and mathematical models for representing all-atom partial charges in each "domain" using a smaller number of "coarse-grained" point charges whose electrostatic potential best mimics that of the original charge distribution. These coarse-grained charges are used to approximate the electrostatic interactions beyond certain threshold cutoff distances, and thus speeding up the calculations.

Surface-based implicit solvent models (GBNSR6)
An accurate description of the solvent environment is essential to estimate free energy of solute-solvent interactions in structural and chemical processes, such as folding or conformational transitions of proteins, DNA, RNA, and polysaccharides, or transport of drugs across biological membranes. The explicit solvent framework, within which the movements of individual solvent molecules are explicitly calculated, suffers from considerable computational costs due to the need to account for the very large number of solvent degree of freedom. An alternative is the implicit solvent framework that is based on representing solvent as a continuous medium instead of individual explicit molecules. My research goal in this project is developing methods to further speed up the computations of implicit solvent models. We developed a surface-based GB model, called GBNSR6, that is available in Amber2015 and higher versions.


Check out my Goolge Scholar for an updated list of publications.

Email: izadi