The thesis
India's diversity is an underexplored asset for discovery.
Genomic-grade data, generated at the point of care and linked back to clinical outcomes, becomes something rare: a population-scale, deeply phenotyped resource. That is what we're building toward drug discovery.
From a bedside sample to a validated, de-risked drug target.
Why it matters
Most genomes studied so far look the same.
The reference panels and polygenic scores the field runs on were built overwhelmingly on European-ancestry cohorts. South Asian genomes, a quarter of humanity, remain thinly represented, which means real biology goes unseen and risk models travel badly.
A representation gap
South Asians are vastly under-counted in public genomic datasets, despite enormous allelic and phenotypic diversity.
Scores that don't transfer
Polygenic risk trained on one ancestry loses accuracy in another, care decisions inherit the bias unless the data is rebuilt locally.
Biology left on the table
Founder effects, consanguinity and population structure surface variants and gene–disease links that homogeneous cohorts can never reveal.
Why India, specifically
One of the fastest places on Earth to find a new target.
This isn't sentiment, it's population genetics. The way India is structured makes causal variants easier to find, and drug targets easier to validate, than almost anywhere else. Three facts do the work.
1 in 6
humans is Indian, yet South Asians are a low-single-digit share of the world's genome-wide studies.
4,600+
endogamous communities, many with founder effects stronger than Ashkenazi Jews or Finns.
2×
more likely a drug target clears trials when it carries human-genetic support.
PCSK9
an entire LDL-lowering drug class that began with people born carrying the gene switched off.
Founder effects, at scale
Centuries of endogamy concentrated rare variants inside communities. What takes a million Europeans to detect can surface in a few thousand genomes here, the same advantage that made Finland and Iceland engines of gene discovery, multiplied across thousands of populations.
Nakatsuka et al., Nature Genetics 2017
Human knockouts
Consanguinity produces people naturally missing both copies of a gene. They show (in a living human, not a mouse) what happens when that gene, and a drug that blocks it, is switched off. South-Asian cohorts have already surfaced thousands of these natural experiments.
Saleheen et al., Nature 2017 (PROMIS) · Genes & Health
Validation that de-risks pharma
The industry now prizes targets with genetic proof, because that proof roughly doubles the odds of surviving the clinic. India can generate it for diseases and ancestries the existing data simply doesn't cover.
Nelson et al., Nature Genetics 2015 · Ochoa et al., 2022
These figures reference published population-genetics and drug-discovery literature, not Wellytics data. They describe the opportunity; the cohort that realises it is what we're building, patient by patient, at the point of care.
The thesis, in motion
Four moves, one compounding loop.
From a bedside sample to a validated, de-risked drug target, and why India is the place to run that loop.
The flywheel
Generated at care. Linked to outcomes. Compounding.
Because the sequencing runs on the same platform, Health Hub, as the clinical agents, every assay ties back to a real, longitudinal clinical record, not a one-off research sample. Every patient makes the resource more valuable.
deepens the resource.
At the point of care
Testing ordered in the clinic, on Health Hub.
Sequenced & QC'd
Joint-called, annotated, research-grade.
Linked in FHIR
Sequence tied to phenotype and outcome.
A deepening cohort
Every patient makes the resource more valuable.
What it unlocks
From a population resource to a discovery engine.
- 01
Target discovery & validation
Gene-burden and association studies on a deeply phenotyped, ancestrally distinct cohort, surfacing and de-risking targets that European-only data would miss.
- 02
Ancestry-aware risk models
Ancestry-aware polygenic scores, applied and calibrated in Indian populations, so risk stratification holds better at the bedside.
- 03
Trial enrichment & recruitment
Genotype-linked clinical records let partners find the right patients faster and design studies around real population structure.
- 04
Biomarker & pharmacogenomics
Linking variants to longitudinal outcomes opens response and adverse-event biomarkers in an under-studied population.
From raw sequence to a druggable target
An agent for every step of discovery.
The analysis layer, turning linked clinico-genomic data into ranked targets, classified variants, longitudinal evidence and research-ready cohorts. Pick one below to watch it work. Molecular testing assays live on the Testing page.
Target Discovery
Runs GWAS and gene-burden analysis across the cohort to surface druggable targets, ranked and annotated.
Variant Interpretation
Classifies variants to ACMG / AMP criteria with evidence, ready for clinical or research sign-off.
RWE Linkage
Links genomes to real-world clinical outcomes, turning sequence into longitudinal evidence.
Cohort Builder
Builds OMOP / HPO-structured, phenotyped cohorts from linked clinico-genomic records.
Who partners with us
Built for the people turning sequence into medicine.
Pharma & biotech
Access to a differentiated cohort for target discovery, validation and trial design, under governance you can stand behind.
Academic & research
Population-scale data and the infrastructure to reason over clinical and genomic records together, for collaborative science.
Public health & consortia
National-scale screening and population-genomics programmes, with on-prem deployment where sovereignty is required.
On the same standards as the clinic
Consent, governance and security are not an afterthought.
The discovery resource is built on the same compliance backbone as the clinical agents, patient consent, de-identification and access controls first, with data residency you control.
HIPAA
GDPR
SOC 2 Type II
ISO 27001
FHIR R4
Build the discovery cohort with us.
Whether you're scoping a target programme, a risk-model collaboration or a population study, let's talk about what the data can do.