Final Schedule

Invited Talks

Machine Learning Approaches in Drug Discovery

The paradigm of drug discovery is undergoing a transformative shift, moving from classical structure-based approximations to data-driven predictive intelligence. This talk elucidates the synergy between fundamental molecular modelling and emerging machine learning (ML) methodologies in accelerating therapeutic development.

Key focus is placed on the development and application of indigenous platforms, such as the Molecular Property Diagnostic Suite (MPDS), which integrates massive chemical libraries with disease-specific workflows. Special emphasis is given to downstream challenges, demonstrating how ML classifiers can effectively predict ADMET profiles (e.g., blood-brain barrier permeability) and assess the probability of clinical trial attrition. The session concludes with a perspective on the future challenges of data quality and the interpretability of AI models in medicinal chemistry.

Quantum Toxicology: De-Risking Paradigm for Predictive Chemical Risk Assessment

The exponential expansion of chemical libraries within agrochemical and pharmaceutical R&D necessitates integrative approaches to navigate chemical space and effectively prevent and address associated toxicities. Predictive cheminformatics presents an efficient, cost-effective means of de-risking chemical toxicity through the combination of computational modelling, experimental datasets, and AI/ML-driven analytics. Methods like QSAR modelling facilitate structure-based predictions of genotoxicity, carcinogenicity, and off-target interactions at reduced dependence on in vivo screening. Complementary systems-level approaches offer a mechanistic understanding of chemical–biological interactions, determining molecular initiating events critical to human health and environmental safety. Emerging paradigms in quantum toxicology enable sub-molecular level prediction of reactivity patterns, enhancing mechanistic clarity in hazard identification. Investigative toxicology frameworks complement in silico, in vitro, and in vivo knowledge, further integrating within the 3Rs to limit animal use. This talk will address emerging research trends in computational approaches for food and chemical risk assessment, antimicrobial resistance, and off-target toxicity of small molecules and drugs. Key aspects related to predictive cheminformatics, quantum toxicology, and AI/ML-driven modelling will be discussed, with examples spanning pharmaceuticals, agrochemicals, and environmental health.

EVANS: An Integrated QSPR Approach for Pharmacokinetics, Pharmacodynamics, and Permeability Modeling

Various tools are available to evaluate individual attributes involved in drug development; however, their fragmented nature often limits routine and integrated application. In addition, many of these tools require complex and resource-intensive inputs, such as knowledge of the bioactive conformation, molecular alignment strategies, and molecular dynamics or quantum mechanical calculations to derive  relevant properties. These limitations motivated the development of a unified approach for analyzing pharmacodynamics (PD), pharmacokinetics (PK), and ADMET characteristics of compounds based on the principles of quantitative structure–property relationships (QSPR). This presentation describes the EVANS formalism, its underlying methodology, and its applications in the evaluation of PD, PK, ADMET properties, and skin permeability.

QSAR and Molecular Modeling in Drug Design and Safety Assessment: From Real Case Studies to AI-Driven Predictive Decision-Making

This lecture presents an integrated view of QSAR and molecular modeling as practical decision-making tools in drug design and environmental health sciences. Using real research case studies—including efflux pump inhibitors, hERG toxicity prediction, and ADME modeling - the talk illustrates how ligand-based QSAR and structure-based approaches complement each other in modern computational drug discovery. Fundamental QSAR concepts are briefly introduced to support young researchers, followed by a forward-looking discussion on AI-driven QSAR, multitask learning, and data-centric modeling strategies. The lecture aims to bridge theoretical foundations, real-world applications, and future perspectives in predictive modeling for safer drugs and chemicals.

Molecular Alchemist - Multidimensional AI strategies for therapeutic solutions, interactions and targeted interventions

In the evolving landscape of pharmaceutical innovation, the research aims to adopt transformative approaches to drug discovery by developing an intelligent, adaptive machine learning ecosystem. The aim is to deconstruct traditional drug development paradigms by integrating advanced computational strategies to navigate complex biological landscapes, generating novel molecular candidates, predicting drug-target interactions, toxicity profiling of molecules and predicting optimal treatment strategies. The focus is on transforming drug discovery from a serendipitous trial-and-error process to a precisely engineered, data-driven scientific approach that accelerates therapeutic development while minimizing experimental iterations and costs.

Integrated structure and machine learning guided Identification of novel scaffolds of therapeutic relevance 

Traditional drug discovery is a notoriously laborious and costly process with pre-clinical stage itself typically taking 3-6 years and costing hundreds of millions dollars. 

ML/DL guided novel workflows augmented with structure- based approaches have shown tremendous promise in harnessing fast evolving computational resources in conjunction with robust programming environment and plenty of useful data and this will further help in extracting knowledge related to molecular structure and activity data to design novel therapeutically relevant molecules against specific targets. This presents a much needed requirement towards providing efficient solutions to drug discovery problems.

A general overview of the role of bio/chem-informatics resources that are of growing importance to drug development process and more specifically integrated structure and machine learning guided Identification of novel scaffolds of therapeutic relevance will be discussed.  

3D QSAR- The basics

QSAR has been a vibrant tool for drug design from decades ago and attempts to quantify the biological activity of the synthesized compounds corresponding to their physicochemical parameters.

Further research has led to 3D QSAR which takes into account the three dimensional structure of the molecules and the receptors, if any. Our lecture will be based upon the basics of 3D QSAR so that the traditionalists can have an understanding of the methods and tools of this process and can intelligently apply 3D QSAR wherever possible. It is hoped that this lecture shall enlighten the audience with a lucid description of the basics of 3D QSAR.

Return to Home