Digital Drug Discovery Company Wins A.I. Competition

(Markus Spiske, Unsplash)

1 May 2018. A company using artificial intelligence algorithms and big data analytics to discover drug targets from RNA splicing errors is the winner of a competition among start-up enterprises for innovative uses of A.I. technology. Envisagenics Inc. in New York is receiving a $1 million investment and additional in-kind credits from M12 — that up to 30 April was known as Microsoft Ventures — and Madrona Venture Group in Seattle.

Envisagenics is a 4 year-old company, spun off from Cold Spring Harbor Laboratory, a biomedical research institute on Long Island, New York. The company identifies therapeutic targets from errors in RNA splicing, a process where the normal progression from genetic code to functioning protein is disrupted. These disruptions are caused by complex interactions among messenger RNA that transmits coding instructions to cells and various early-stage proteins.When built-in regulatory steps fail to fix the disruptions, splicing errors can occur.

While errors account for a small proportion of RNA splices, Envisagenics says they can be traced to at least 370 genetic diseases, including spinal muscular atrophy, Huntington’s disease, and acute myeloid leukemia.The company cites data showing about 15 percent of all diseases are caused by disrupted splicing, accounting for about half of rare genetic disorders.

Envisagenics’s technology, known as SpliceCore, is designed to identify RNA-based therapeutics that can correct splicing errors. The company says its algorithms use machine learning trained from public RNA sequencing databases and its own data stores, from which it develops a predictive model. That model, residing in the cloud, is used to evaluate a patient’s RNA sequences to identify splicing errors, as well as associated drug targets and biomarkers.

The company says SpliceCore is faster and more accurate than other RNA drug discovery processes. Each splice event, says Envisagenics, is treated as a separate statistical inference problem that reduces the complexity transcription analysis to focus on drug targets and biomarkers. The process that begins with some 5 million splicing events is winnowed down with RNA mapping, experimentation, and statistical modeling to result in a single feasible therapeutic target.

Envisagenics took part in the latest Innovate.AI challenge for start-up enterprises harnessing machine learning and artificial intelligence. The company took part in the North American competition, held at the same time as similar competitions in Europe and Israel, and came away with the top prize of a $1 million investment. Envisagenics is also eligible for Microsoft Azure cloud computing credits of $500,000 a year for 3 years, as well as 100 Microsoft Office licenses for 1 year.

The company was founded in 2014 by Maria Luisa Pineda, a Cold Spring Harbor PhD recipient and Martin Akerman, a postdoctoral researcher at Cold Spring Harbor. Pineda and Akerman are now Envisagenics’ CEO and chief technology officer respectively.

More from Science & Enterprise:

• Platform to Integrate Robotics, Synthetic Biology, A.I.
• Self-Service A.I. Offered for Pathologists
• A.I./Image Analysis Diagnose Eye, Lung Diseases
• Brain Wave Data Harnessed for Open-Source Brain Model
• Microsoft Providing A.I. for Big Data Cancer Research

*     *     *