One discovery engine. Seven ways to use it.
A single labelled cohort — taxonomy, metabolomics, functional profiling — can be interrogated for several distinct clinical questions at once. Below, each application is explained in plain terms, then grounded in a real MANTRA finding so you can see exactly what it looks like in practice. Each finding is also tagged by what kind of result it is: a known clinical distinction MANTRA recovered independently, a previously unreported relationship it found, a mechanism now ready for India-specific validation, or — for the last item below — a platform capability rather than a single finding. For the same clinical findings organised by therapeutic area instead of function, see Evidence.
Telling apart diseases that look alike on presentation
Some conditions overlap enough in symptoms and standard tests that getting the subtype wrong delays the right treatment. MANTRA builds a multi-biomarker signature that separates them from a single sample.
Ulcerative Colitis vs. Crohn's Disease — overlapping presentation otherwise complicates diagnosis and treatment selection. MANTRA distinguishes the two subtypes from gut bacteria and metabolite signals alone, validated against the HMP2/IBDMDB and Franzosa/PRISM cohorts.
Objectively grouping patients who will respond differently to the same drug
Gut microbes directly metabolise many chemotherapeutic drugs, changing both how well they work and how toxic they are. MANTRA classifies patients into clinically meaningful groups before treatment starts.
Treatment response stratification — MANTRA is designed to separate colorectal cancer patients likely to respond to standard chemotherapy from those who won't, at the individual patient level, before treatment begins.
Seeing a complication coming, months before it's clinically visible
Standard tools confirm inflammation is present, not which direction a patient is heading. MANTRA identifies early biological signatures of progression that clinicians currently have no way to anticipate.
Early complication warning in IBD — MANTRA is designed to flag patients at higher risk of disease complications before they become visible clinically, so treatment can be escalated earlier rather than after the fact.
Flagging who will suffer severe treatment toxicity, before treatment begins
Genetics alone misses a real share of at-risk patients — especially where Indian-specific variant frequencies differ from the Western populations standard panels were built on. MANTRA adds the microbiome layer standard testing doesn't look at.
Chemotherapy toxicity risk — a large share of patients on some widely used chemotherapies develop significant toxicity. Combining genetic and microbiome data into a single risk score, run once before the first cycle, is designed to flag this before it happens — using population-specific data standard panels don't screen for.
Uncovering the disease-driving pathway conventional analysis missed
Not every finding needs to be a biomarker panel — sometimes the most valuable output is a testable mechanism: which enzyme, which metabolite, which downstream effect. That's a lead for a pharma partner, not just a diagnostic.
Shared metabolic signals in IBD — MANTRA independently identified metabolic signals common to both Ulcerative Colitis and Crohn's Disease that were missed in prior published analysis of the same public patient data, pointing to shared inflammatory mechanisms that may be druggable.
Turning a protective or causal microbial signal into something to build, not just measure
Not every microbial signal is best used as a marker. When MANTRA identifies an organism that's causally protective or causally driving disease — not just correlated with it — that organism itself becomes a candidate: a strain to develop as a live biotherapeutic, or a target for a pharma partner to drug.
A protective organism identified alongside one of MANTRA's diagnostic findings — the same organism used as a point-of-care test candidate is also a live-biotherapeutic candidate in its own right: something to deliver, not just detect. Full details are under active patent review.
Will a candidate strain actually survive in the target gut?
Gut microbes aren't independent — they compete for the same nutrients, cooperate by cross-feeding metabolites to each other, and depend on keystone species that anchor the whole community's structure. Introducing a new strain into that environment doesn't guarantee it survives: whether it engrafts depends on whether it competes with or complements the organisms already there, and whether the niche it needs is already occupied.
MANTRA's output isn't a species list, it's a network — nodes are microbes, metabolites and enzymes, edges show ecological relationships. That answers the engraftment question before a single trial: which organisms would compete with a candidate strain for the same resources, which ones it could cross-feed from (or would need to), and which keystone species its introduction would displace or depend on.