Enveda Biosciences, a biotechnology company discovering new medicines from natural sources, released the details of one of its foundational AI models, MS2Mol, in a pre-print posted on ChemRxiv. MS2Mol is designed to predict the structure of metabolites, which are the building blocks and breakdown products of the cell. Despite their essential role in all cell processes, it is estimated that less than 1% of all naturally-occurring metabolites are known to science. The ability of MS2Mol to rapidly predict the structure of previously uncharacterized metabolites without lengthy laboratory experimentation enables their prioritization as potential drug candidates and expands our knowledge of the natural world.
“Metabolites have a long and successful history as the basis for impactful drugs including aspirin, taxol, metformin, artemisinin, and statins. This is particularly impressive given that we, as scientists, have barely scratched the surface of natural metabolite diversity. With MS2Mol integrated into our platform, we can tap evolutionary chemical intelligence for the next generation of powerful medicines at scale,” said Viswa Colluru, Ph.D., founder and CEO of Enveda.
Enveda’s proprietary platform solves the long-standing obstacles in natural product drug development including active molecule identification, property and structure prioritization, amenability to medicinal chemistry, and large-scale material access. The company recently closed its Series B1 round and will progress multiple platform-derived molecules to the clinic in 2023 and 2024 across inflammation, fibrosis, and neurosensory indications.
“Metabolite identification used to be a process that was time-consuming, prone to failure, and required highly specialized expertise. MS2Mol takes the most easily accessible – but extremely cryptic – form of data on metabolites, the mass spectrum, and translates it into a language that scientists can use: the chemical structure. Solving this translation problem with AI puts the most useful information in the hands of drug hunters at massive scale,” said David Healey, Ph.D., VP of Data Science at Enveda and senior author of the pre-print.
“While other companies use AI to predict what you want to buy, we use AI to discover what humanity needs to know,” said Tom Butler, Ph.D., VP of Machine Learning at Enveda and first author of pre-print.
“Unlocking bioactive chemistry honed by billions of years of evolution for modern drug discovery has led us to discover a slate of exciting candidate medicines at a remarkable pace,” said Sotirios Karathanasis, Ph.D., CSO at Enveda. “We look forward to modification of disease pathophysiology by our medicines in the clinic and redefinition of the concept of target undruggability with the Enveda platform.”
With the creation of MS2Mol, Enveda continues to deliver field-changing technology for the discovery and utilization of natural metabolites to drive novel therapeutic development.
The paper can be found on http://www.chemrxiv.org under “MS2Mol: A transformer model for illuminating dark chemical space from mass spectra”
About Enveda Biosciences
Enveda Biosciences is a biotechnology company building the first high-resolution chemical map of the natural world to tackle the toughest problems in drug discovery. Enveda’s platform is the world’s most advanced drug discovery search engine from the expanse of nature’s unknown chemistry, building on years of cutting-edge advancements at the intersection of metabolomics and machine learning. Complementing its breakthrough technology, Enveda’s team includes seasoned drug hunters with decades of experience in the pharmaceutical industry working with preeminent data scientists. For more information on Enveda, visit envedabio.com.
View source version on businesswire.com: https://www.businesswire.com/news/home/20230621191398/en/