Researchers from the Institute of Science Tokyo and Alivexis, Inc. have unveiled PairMap, a groundbreaking computing method that improves the accuracy of predicting the free energy of binding medicinal molecules. This significant advancement, recently published in the Journal of Chemical Information and Modeling, has the potential to reduce costs and hasten the development process of new drugs.
Free binding energy is crucial in the creation of new drugs as it helps forecast how effectively a molecule will interact with its target. However, conventional methods like relative binding of free energy (RBFEP) face challenges in modeling extensive chemical transformations or complex restructuring of molecules.
PairMap tackles this issue by introducing an intermediate compound system, which establishes a step-by-step pathway for converting between the molecules under study. This methodology reduces calculation errors, speeds up convergence, and cuts down computational costs.
Tests conducted on standard data sets demonstrated that PairMap outperformed traditional methods significantly. The average absolute error in predicting binding energy dropped from 1.70 to 0.93 kcal/mol. Particularly noteworthy was PairMap’s exceptional accuracy in complex chemical transformations, surpassing existing methods like absolute binding of free energy and RBFEP.
Professor Masahito Oue, one of the study’s authors, remarked, “PairMap has the potential to revolutionize the drug development process. By leveraging carefully selected intermediate compounds and thermodynamic cycles, we have achieved unprecedented prediction accuracy.”
The innovative PairMap algorithm enables in-depth exploration of the chemical space, paving the way for the creation of new, more effective drugs. Future plans for research include expanding the algorithm’s capabilities to accommodate compounds with significant charge changes, making the method even more versatile.