STATUSAGREEMENT IN PROGRESS SCOPEPoC A · CMD FORMULATION SCREEN DATAPUBLIC SOURCES ONLY UPDATED[DATE]
CLASSICAL PHENOTYPE × COMPUTATIONAL SCALE

Which traditional signals survive computation?

Ayurnidaan brings structured classical knowledge and clinical phenotyping. AI Ops Pros brings pipelines, compute and a research scholar network. Together we test which traditional principles produce reproducible biological signal — starting with public data only.

SCROLL
01 — THE QUESTION

Precision medicine needs ways to group people. Ayurveda already has one.

The problem

Modern precision medicine needs frameworks that group individuals by biologically meaningful phenotypes — and reactive care only acts once disease has already manifested.

The premise

Ayurveda operates a centuries-old stratification system built for exactly that purpose: Prakriti as a stable baseline, Vikriti as deviation from it.

The question

Do these frameworks contain reproducible biological signals that can be validated with modern genomics and machine learning?

CORE FOCUS

Cardio-metabolic disease and lifestyle disorders, intercepted at the stratification and prevention stage: understand the individual, identify biological risk, personalize the intervention, monitor longitudinally.

02 — EVIDENCE

What is established, and what is still a hypothesis

The line between the two is the whole point of the collaboration. Everything on the left has peer-reviewed support; everything on the right is what we would be testing.

Established

India's CSIR-IGIB / TRISUTRA consortium has led 15+ years of peer-reviewed Ayurgenomics research. Healthy individuals of contrasting extreme Prakriti types show statistically significant differences in biochemical, haematological and gene-expression parameters, including EGLN1 variation linked to high-altitude adaptation.

Machine learning recovers Prakriti categories from structured phenotypic traits alone (Tiwari et al., PLOS ONE, 2017).

Multiple independent meta-analyses across 100+ randomized trials report that curcumin significantly reduces fasting blood glucose, HbA1c and LDL in type 2 diabetes. Clinical trials (Biswal et al., 2013) report reduced chemotherapy-induced fatigue with Ashwagandha in breast cancer patients.

The infrastructure exists: IMPPAT, a curated public database of roughly 4,000 Indian medicinal plants and 18,000 structure-annotated phytochemicals, built for network-pharmacology screening.

Hypothesis — what we would test

Can scalable phenotypic data map consistently onto reproducible biomarker and clinical signatures?

Does Prakriti-stratified lifestyle intervention, with specific botanical adjuncts, modulate validated inflammatory and arterial-stiffness markers?

Does longitudinal Vikriti tracking correlate with measurable physiological change — heart-rate variability, continuous glucose?

Can multi-target mapping of fatigue and HPA-axis pathways using IMPPAT phytochemicals generate candidate mechanisms worth testing? Herb–drug interaction screening is a prerequisite, not an afterthought.

The execution gap is the honest part: computational predictions generate hypotheses. They still require experimental and clinical validation.

03 — THE BRIDGE

Domain data meets computational scale

AYURNIDAAN CONTRIBUTES
  • Classical literature structured for computational use.
  • Prakriti and Vikriti phenotyping frameworks, with assessment tooling (PAQ-25).
  • BAMS-qualified clinical expertise and hypothesis generation.
  • Clinical parameters, symptoms, lifestyle patterns and health history, tracked longitudinally.
  • Institutional grounding: BITS Pilani / Rakesh Kapoor Innovation Center, FITT IIT Delhi.
AI OPS PROS CONTRIBUTES
  • AI, ML and LLM engineering, with infrastructure built for scale.
  • Sequencing analysis across DNA, RNA and bisulfite pipelines.
  • Network pharmacology and target-prediction workflows.
  • A research scholar network, mentored and reviewed by a PhD lead.
  • The Sutra Loop engine: research papers turned into structured reports and reusable workflows.
MUTUAL BENEFIT

Ayurnidaan gains robust ML infrastructure and pipeline execution for a domain that has had little of either.

AI Ops Pros gains access to a novel, source-traceable research pipeline and a scientifically serious problem to work on.

04 — PROOF OF CONCEPT

PoC A: a zero-risk computational screen

Can we computationally screen one specific classical cardio-metabolic formulation against known CMD gene networks? Public data, weeks not quarters, and a result that tests the working relationship before anything scales.

INPUT · PUBLIC DATA

Formulation and phytochemicals

One classical cardio-metabolic formulation, expanded into its constituent phytochemicals through IMPPAT 2.0.

COMPUTE · AI OPS PROS

Targets, network, hubs

SwissTargetPrediction for candidate targets, STRING for protein interaction network construction, Cytoscape for hub-gene analysis.

OUTPUT · JOINT

Ranked, benchmarked targets

A ranked list of candidate molecular targets and pathways, benchmarked against known CMD gene networks — publishable and IP-eligible.

The rules. Public datasets only — IMPPAT, GEO, STRING. No proprietary patient data exchanged. No IP entanglement at this stage.

The cost. The target is an analysis that runs for a few dollars rather than a few thousand, so cost stops being the reason a question goes untested.

The honest limit. A ranked target list is a hypothesis generator. Experimental and clinical validation still decides what is true.

05 — DATA AND HANDLING

Built to respect where the data lives

Health data carries obligations. The working arrangement is designed so that those obligations shape the pipeline rather than being bolted on afterwards.

DE-IDENTIFICATION

Codes, not identities

Patient identifiers are stripped before anything moves. Internal codes are enough for the analysis; only minimal metadata is required.

RESIDENCY

India-based analysis, if needed

If data must stay within India, India-based scholars run the analysis and the US-based leads restrict themselves to high-level review.

SEQUENCING

RNA over DNA, for tracking change

Germline and somatic variants stay largely fixed, so before-and-after studies gain more from RNA sequencing — or methylation — where the signal is dynamic. DNA, RNA and bisulfite pipelines are all supported.

INTAKE

Raw files accepted as they come

Illumina BaseSpace exports or FASTQ both work. A standard operating procedure for formats and transfer is provided before the first handoff.

GUARDRAIL

Ayurveda is never positioned as a cancer cure or as a replacement for standard oncology regimens. Supportive-care claims stay inside what the trial evidence supports.

06 — NEXT 90 DAYS

Collaboration, measured by what ships

No fees during this phase. What we ask in return is a testimonial, joint publications, and introductions to your network once the work has proven itself.

THIS WEEK

Profiles and paperwork

Scholar résumés and LinkedIn profiles shared by Drive. Ayurnidaan sends a brief description of the data on hand. The agreement moves in parallel.

WEEKS 2–4

Team prepared, pipeline drafted

Scholars are briefed on the domain and the workflow is architected, so analysis can begin the day the agreement is signed.

MONTH 2

PoC A executed

The formulation screen runs end to end on public data, with the method documented as it goes.

MONTH 3

Result, review, decision

Ranked targets reviewed jointly, a write-up drafted for publication, and a decision on whether to scale to patient-data work.

07 — WHO IS DOING THE WORK

Clinicians, engineers, scholars

[PHOTO]

Dr. Nikhil Khatana

FOUNDER & DIRECTOR · AYURNIDAAN

BAMS. Bridges classical literature and clinical preventive care.

[PHOTO]

Prof. Hare Krishna Mohanta

CO-FOUNDER · AYURNIDAAN

24+ years in academia and research, translating complex systems into computational models.

[PHOTO]

Pawan Bhat

SCIENTIFIC LEAD · AI OPS PROS

PhD. Genomics and transcriptomics analysis; architects the workflows and reviews every result.

[PHOTO]

Yuva Raju Nadimpalle

ENGINEERING LEAD · AI OPS PROS

~20 years in software. Builds the pipelines, the infrastructure and the Sutra Loop engine behind the team.

10research scholars onboarded, with résumés and profiles shared
8 in Indiaplus US-based scholars, so data residency has a clean answer
~100scholars targeted by 2027
THE NEXT STEP

Align on PoC A and validate the pipeline.

Send a short description of the data you have on hand, and we will have the team prepared before the paperwork is finished.