Accelerating Molecular Design with Machine Learning and AI

Progress in autonomous and high-throughput experimentation and the ongoing maturation of machine learning (ML) and generative AI has led to a drastic increase in the pace of chemical and materials research. In this talk, we will overview two directions that emphasize how machine learning and experiments/simulations can be combined.

First, we will present a generative AI method for de-novo design of molecular compounds with desired biological properties. The general workflow of the proposed method integrates two deep neural networks – generative and predictive – that are initially trained separately but then trained jointly to generate novel chemical structures with desired properties. In the proof-of-concept study, we have employed this integrative strategy to design chemical libraries biased toward compounds with specific ranges of physical properties (like solubility and hydrophobicity) as well as to develop novel kinase inhibitors of JAK2, EGFR, and CDK1. This new approach can find general use for generating targeted chemical libraries optimized for a single desired property or polypharmacology.

See link below for the full abstract.

A Radical Solution for C(sp3)–C(sp3) Bond Formation during the Biosynthesis of Macrocyclic Membrane Lipids

Archaea synthesize isoprenoid-based ether-linked membrane lipids, which enable them to withstand extreme environmental conditions, such as high temperatures, high salinity, and low or high pH values. In some archaea, such as Methanocaldococcus jannaschii, these lipids are further modified by forming carbon–carbon bonds between the termini of two lipid tails within one glycerophospholipid to generate the macrocyclic archaeol or forming two carbon–carbon bonds between the termini of two lipid tails from two glycerophospholipids to generate the macrocycle glycerol dibiphytanyl glycerol tetraether (GDGT). GDGT contains two 40-carbon lipid chains (biphytanyl chains) that span both leaflets of the membrane, providing enhanced stability to extreme conditions. How these specialized lipids are formed has puzzled scientists for decades. The reaction necessitates coupling two completely inert sp3-hybridized carbon centers, which has not been observed in nature. Here we use X-ray crystallography, high-resolution mass spectrometry, chemical synthesis, and biochemical analyses to show that the gene product of mj0619 from M. jannaschii, which encodes a radical S-adenosylmethionine enzyme, is responsible for biphytanyl chain formation during synthesis of both the macrocyclic archaeol and GDGT membrane lipids.