Can it take just a few minutes to calculate how proteins interact with drugs?

A research team from the Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences has developed a novel computational method that can accurately describe how proteins interact with molecules of potential drugs and do so in a mere few tens of minutes. This new quantum-mechanical scoring function from IOCB Prague can thus markedly expedite the search for new drugs. A scientific article on the topic has been published in the journal Nature Communications.

The study demonstrates that this is the first universally applicable method of its kind. Computational experts from IOCB Prague tested it on ten proteins of different levels of structural complexity, each binding a large variety of small molecules (usually referred to as ligands). They then compared their results not only with those of other corresponding methods, but also with findings of laboratory experiments, and both comparisons turned out very favourably.

‘Of course, we are not the only ones working on this. There are several such methods. Usually, however, their speed is offset by low accuracy whereas more accurate calculations can take several days. Our methods are unique in that they can process information about large molecular systems within tens of minutes while retaining the benefits of much more demanding quantum-mechanical calculations,’ explains Jan Řezáč, corresponding author of the article from the Non-Covalent Interactions group led by Prof. Pavel Hobza.

Experts from this group have been studying intermolecular interactions for a long time. In this research they focus mainly on biomolecules, and the results of their work directly bear on the computer-aided design of drugs. The reason is that when scientists work toward a new drug, they often look for molecules that bind strongly to a particular protein. Identifying them, however, is akin to finding needles in a haystack, as large numbers of molecules have to be tested to set apart those that show promise. This considerably slows down the discovery of medicinal substances and makes it more expensive. By predicting the strength of protein–ligand binding, and thus singling out molecules that best satisfy a defined set of criteria, computational chemists spare the work of experimenters, which, in turn, significantly accelerates drug discovery.

The potential of this research is so great that it has received support from the commercial sector, and experts from IOCB Prague have been cooperating with one of the world's largest pharmaceutical companies for some time.

Original article: Pecina, A., Fanfrlík, J., Lepšík, M., Řezáč J. SQM2.20: Semiempirical quantum-mechanical scoring function yields DFT-quality protein–ligand binding affinity predictions in minutes. Nat. Commun. 15, 1127 (2024). https://doi.org/10.1038/s41467-024-45431-8