IITD researchers design tool for synthesizing molecules

In what can be a boon for the pharma and biochemical industry, researchers of IIT Delhi in collaboration with IIT Madras have designed an algorithm that can help synthesize new molecules.

The software tool that has been named as ReactionMiner can help tell the best pathway for designing a molecule.

“The ability to predict pathways for biosynthesis of metabolites is very important in metabolic engineering. It is possible to mine the repertoire of biochemical transformations from reaction databases, and apply the knowledge to predict reactions to synthesize new molecules. However, this usually involves a careful understanding of the mechanism and the knowledge of the exact bonds being created and broken,” says lead researcher Sayan Ranu, assistant professor in the department of computer science of IIT Delhi.

The research was recently published in the journal Bioinformatics.

Emphasizing on the need for a method to rapidly predict reactions for synthesizing new molecules, which relies only on the structures of the molecules, without demanding additional information such as thermodynamics or hand-curated reactant mapping, which are often hard to obtain accurately, Ranu said: This tool that is fully automated can help in better predictions. Our approach scales well, even to databases with >100 000 reactions.

The next goal he says is to make a server based web application where people can easily use the tool. Elucidating on how this differed from other tools he said: “Besides better quality this gives an interpretable result. It can essentially tell why we are predicting. Other tools are less transparent.”

The Mechanism

The researchers have described a robust method based on subgraph mining, to predict a series of biochemical transformations, which can convert between two (even previously unseen) molecules. They first described a reliable method based on subgraph edit distance to map reactants and products, using only their chemical structures. Having mapped reactants and products, they identified the reaction centre and its neighbourhood, the reaction signature, and stored this in a reaction rule network. This novel representation enabled them to rapidly predict pathways, even between previously unseen molecules. The researchers demonstrated this ability by predicting pathways to molecules not present in the KEGG database. They also proposed a heuristic that predominantly recovered natural biosynthetic pathways from amongst hundreds of possible alternatives, through a directed search of the reaction rule network, enabling one to provide a reliable ranking of the different pathways.