Predicting Lead Structure(s) for Molecular Library Screening Using Molecular Docking and Quantum Crystallography

Posted: 11 Feb 2020

See all articles by Suman Kumar Mandal

Suman Kumar Mandal

Shiv Nadar University - Department of Chemistry

Parthapratim Munshi

Shiv Nadar University - Department of Chemistry

Date Written: February 6, 2020

Abstract

Optimization of lead structure is important for drug discovery studies and requires improvements via chemical development, guided by available ligand SAR and 2D/3D similarities. Accurate prediction of experimental ligand binding affinity using MD simulations, Monte Carlo etc. or protein-ligand interaction energies using MM/PBSA or MM/GBSA can help accelerate the process, but these approaches are computationally expensive and impractical for routine evaluations like molecular library screening. Molecular docking and scoring functions are very prominent tools for such studies being computationally inexpensive. Although these scoring functions are widely used for comparing ligand potencies, none of the scoring function can be labelled ‘reliable’ and their accuracy of prediction can fall anywhere between 0-93%. Introducing a reference ligand (lead structure) can increase the accuracy many folds for these scoring functions. The lead structure depends on the availability of protein-ligand complex structure; hence the concern remains unanswered when there is no complex structure available. The quantum crystallographic technique Kernel Energy Method (KEM), is a computationally inexpensive fragment-based approach that accurately calculate the interaction energies of macromolecules using their crystallographic coordinates. In this study we show a novel approach to accurately predict the lead structure(s) for proteins using scoring function (GoldScore) along with quantum crystallography (modified-KEM). Human aldose reductase (hAR) and Cyclin dependent kinase 2 (cdk-2) with resolution 0.66 Å (PDB ID- 1us0) and 2 Å (PDB ID- 1h1s), respectively were chosen and their known ligands selected from the literature were docked into the respective protein structures using GoldScore. Interaction energies of the docked poses with the active site residues are calculated using modified-KEM. Based on the interaction energy of the poses we predicted the lead structure(s) for hAR and cdk-2 and validated the results by comparing the deviation of predicted lead structure from the experimental lead structure (geometry of ligand in protein-ligand complex).

Keywords: Quantum Crystallography, Kernel Energy Method, Lead Structure, Library Screening

Suggested Citation

Mandal, Suman Kumar and Munshi, Parthapratim, Predicting Lead Structure(s) for Molecular Library Screening Using Molecular Docking and Quantum Crystallography (February 6, 2020). Proceedings of International Conference on Drug Discovery (ICDD) 2020, Available at SSRN: https://ssrn.com/abstract=3532904

Suman Kumar Mandal (Contact Author)

Shiv Nadar University - Department of Chemistry ( email )

NH-91, Village- Chithera, Tehsil-Dadri,
Dist. Gautam Buddha Nagar
Gautam Buddha Nagar, Uttar Pradesh 201314
India

Parthapratim Munshi

Shiv Nadar University - Department of Chemistry ( email )

NH-91, Village- Chithera, Tehsil-Dadri,
Dist. Gautam Buddha Nagar, UP
Gautam Buddha Nagar, Uttar Pradesh
India

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
298
PlumX Metrics