Code Development
Codes developed for Bio-computing and Materials modeling
BioFrag : Many of the drug-like properties are related to free energy difference of a ligand in two different molecular environments. For example, the binding affinity of a ligand is related to its free energy difference in biomolecular target and aqueous envrionments. In order to estimate the binding affinities and other drug-like properties, we need to estimate the free energy of ligands to an accuracy of a few kcal/mol. The force-field methods often fail in correctly predicting the relative binding affinity of ligands and a electronic structure theory based method needs to be used. The reason for the improved performance is due to the fact that this method can account for mutual polarization between the protein, ligand subsystems in a complex. However, the use of this approach is limited to a system having a few hundreds of atoms and is difficult to employ it for describing the whole protein-ligand system. I have developed a fragmentation scheme to compute the total interaction energy between the protein-ligand as the sum over contributions from individual ligand-aminoacids. Further, the scheme developed can employ a hydrogen capping method or capping with N-methyl amino and acetyl groups to avoid charge accumulation in the dangling bonds in individual aminoacids after the fragmentation. This is based on the approximation that the total interaction of a collection of molecular fragments can be described as the sum over two body interactions. Eventhough it is often the case, it is not necessarily be followed for all cases. For some systems, the three body contributions can be significant. To respect this aspect, the scheme also can provide the interaction energy as the sum over ligand-dipeptide contributions and this can be used to estimate the contributions due to three body effects. The benchmarking of developed QM fragmentation method to correctly score binding affinity of various ligands towards Alzheimer's related biomolecular targets is in progress.
• BioPFF / PeptPFF: Modeling electronic and magnetic properties of ligands bound to biomolecules or in solvent environments is challenging as we cannot treat the enviroment at the same level of theory as the ligands. As an approximation, the interaction between ligand and environment can be described using an effective Hamiltonian which includes electrostatic, van der Waals interaction between two subsystems. Currently there are many embedding schemes available to describe the subsystem interactions effectively. Among these electrostatic and polarizable electrostatic embedding schemes reliably and effectively describe the heterogeneous molecular environment around a ligand which is described using electronic structure theory method. However, the charges, dipoles and other higher order mutipoles of charge distribution and atom centered polarizabilities have to be computed for the surrounding environment. It is not a problem for the solvent environment as these properties can be computed for individual molecules. However, this is not stright forward for a biological environment such as proteins, DNA, RNA. I have developed a fragmentation scheme where the whole protein can be fragmented into individual aminoacids and these properties can be computed. The charges and polarizabilities computed for the capping hydrogens should be moved to the edge atoms and this way the charges and polarizable force-field for entire bioenvironment can be computed. The importance of using polarizable force-field for describing the biomacromolecular environments for reliable and accurate modeling of optical (one photon and two photon) and magnetic properties of small molecules bound to these biological targets has been discussed in some detail in literature. Currently available polarizable force-fields (AMOEBA) are limited to modeling the structure and dynamical properties. I have developed codes/software to automatically generate poloarzable force-field for proteins, DNA and membranes. The software will make use of the available electronic structure theory software, DALTON2016 and package LoPROP (developed majorly in Scandinavia and with contributions from KTH). The input configuration for the biomacromolecule for which the polarizable force-field needed can be extracted from trajectories from molecular dynamics or hybrid QM/MM molecular dynamics.
• FloppyMC: I have developed (thanks to Prof. S. Yashonath, IISc and Dr. Anil Kumar, NISER for useful discussions) variable shape variable size Monte Carlo code in isothermal- isobaric ensemble to study organic molecular crystals as a function of temperature and pressure. The code includes features to use a flexible molecular model for floppy molecules by explicitly incorporating the low frequency large amplitude modes and so in addition to thermal and pressure effect on the crystal structure, packing, the molecular structure and conformational changes can be studied. The pressure and tenperature dependent studies carried out on a number of flexible organic molecules such as biphenyl, p-terphenyl, stilbene, and 4-vinyl benzoic acid could explain the microscopic origin of a number of phase transitions reported in these molecular crystals based on experimental IR , Raman spectra, X-ray diffraction and calorimetric studies. Possible explanations for the temperature-induced anomalous bond length shrinkage in 4-vinyl benzoic acid have been obtained based on its temperature dependent average molecular geometry and a set of ab initio constrained optimizations (refer to Murugan, JCP, 123, 094508 (2005)) . In molecular crystals of stilbene and adamantane the nature of conformational disorder and molecular geometry has been reported to be dependent on the specific crystallographic site (refer to Murugan, et al., JPCB, 109, 17296 (2005) and Murugan et al., JPCB 109, 12107 (2005)) which is yet to be verified from experiments. Further a number of analysis codes have been developed to study the structure and dynamic properties of molecules in solvent and bio-environments. Possible explanations for the temperature-induced anomalous bond length shrinkage in 4-vinyl benzoic acid have been obtained based on its temperature dependent average molecular geometry and a set of ab initio constrained optimizations (refer to Murugan, JCP, 123, 094508 (2005)). In molecular crystals of stilbene and adamantane the nature of conformational disorder and molecular geometry has been reported to be dependent on the specific crystallographic site (refer to Murugan, et al., JPCB, 109, 17296 (2005) and Murugan et al., JPCB 109, 12107 (2005)) which is yet to be verified from experiments. Further a number of analysis codes have been developed to study the structure and dynamic properties of molecules in solvent and bio-environments. Possible explanations for the temperature-induced anomalous bond length shrinkage in 4-vinyl benzoic acid have been obtained based on its temperature dependent average molecular geometry and a set of ab initio constrained optimizations (refer to Murugan, JCP, 123, 094508 (2005)) . In molecular crystals of stilbene and adamantane the nature of conformational disorder and molecular geometry has been reported to be dependent on the specific crystallographic site (refer to Murugan, et al., JPCB, 109, 17296 (2005) and Murugan et al., JPCB 109, 12107 (2005)) which is yet to be verified from experiments. Further a number of analysis codes have been developed to study the structure and dynamic properties of molecules in solvent and bio-environments.
• ToolsQMMM: I have developed a number of codes / scripts for computing and analyzing one and two photon properties of molecules bound to biomacromolecules like fibril, enzyme, membrane, DNA and so on. The program uses the trajectory files from MD or QM / MM MD and generates input files for further running TD-DFT / MM calculations using DALTON2016.
