SOMA: Research Automation Platform

You give SOMA a pair of concepts, say a lifestyle factor and a medical condition, and it returns the causal chains linking them, each backed by sentences pulled straight from research articles.

That is the core of SOMA, a research automation platform for finding pathways between factors and medical conditions that nobody has documented in one place.

Behind the search sits a database built from open access journals.

SOMA analyzes medical research articles, extracts the key concepts, and identifies causal and associative relationships between them.

Those relationships are organized into a knowledge graph, which is what speeds up the search compared with reading papers one by one.

The sentence-level evidence is the useful part.

You can check each link in a chain against the wording of the source article instead of trusting a summary.

The catch is scope: because the database draws on open access journals, paywalled literature is not part of what SOMA searches.

It works for researchers exploring hypotheses in health and medicine.