Property-Driven Molecular Optimization Framework

This document is a technical report detailing the SynthonForgePlus framework, which facilitates property-driven and multi-objective molecular design. It outlines the complexity of chemical space and the need for goal-directed exploration to efficiently navigate this vast area. The report explains the concept of synthons as reactive building blocks that enhance the exploration process by allowing for scalable and relevant compound generation. It describes the integration of ADMET Predictor for assessing pharmacokinetic endpoints and the use of multicriteria decision analysis (MCDA) techniques, specifically the VIKOR method, for ranking alternatives based on conflicting criteria. The document also presents a use case involving the optimization of RORγ agonists, detailing the setup, objectives, and results achieved through the framework. Overall, the report emphasizes the importance of a structured approach to molecular design in drug discovery, highlighting the iterative nature of the optimization process and the role of user-defined parameters.