A model built for a defined question
A human virtual twin brings biological, behavioral and contextual data into a model that can be updated over time. Its usefulness depends on the question being asked and the evidence supporting that use. An individual-identification model is not, by itself, a disease-detection model.
This article is an overview, not a clinical protocol or a replacement for the methods reported in each publication.
Data and analytical quality
Depending on the program or study, inputs may include metabolomic samples, wearable observations, body measurements, questionnaires and reported environmental or lifestyle events. Not every source is available or appropriate in every setting.
For dried blood spot metabolomics, the published work describes LC-MS analysis, quality controls, normalization and feature selection. Untargeted signals must be distinguished from identified metabolites and clinically validated biomarkers. The number and meaning of the features depend on the method.
Batch effects and repeated samples from the same participant can make a model appear more accurate than it is. BioTwin’s identification study demonstrates why validation must account for these dependencies. A held-out analytical batch tests a different question from a random split of related samples.
What the publications establish
- Individual profiling studies examine whether metabolomic patterns distinguish participants under the reported validation conditions.
- The breast-cancer preprint evaluates classification in a retrospective case-control cohort. It does not establish a population-wide prospective screening benefit.
- The fatigue preprint explores self-reported PEM-like fatigue. It does not validate a diagnostic test for ME/CFS.
- Wearable research evaluates agreement between devices; agreement is not the same as clinical accuracy.
These are distinct tasks. Their results cannot be transferred automatically to cardiology, endocrinology, neurology or mental-health screening. Each proposed indication needs its own appropriate validation.
Research, wellness and clinical access
TwinMe provides non-medical wellness insights. BioTwin’s regulated medical applications are separate and available only in authorized jurisdictions and clinical settings. Current clinical access and its boundaries are described on the oncology page.
A model’s uncertainty, intended population, limitations and potential consequences must be assessed before clinical use. No model replaces clinical judgment or established screening pathways.
Sources and further reading
- Human digital twin technology for individual profiling using LC-MS untargeted metabolomics analysis of dried blood spot samples
- Metabolomic fingerprinting from dried blood spots enables individual identification across 1,257 participants at 94% user-level accuracy
- Metabolomic Profiling of Dried Blood Spots for Breast Cancer Detection: A Multi-Classifier Validation Study in 2,734 Participants
- Exploratory Dried Blood Spot Metabolomics Identifies Pathway-Level Convergence with ME/CFS Biology in a Self-Reported PEM-Like Fatigue Phenotype
- Multi-Year Longitudinal Inter-Device Agreement Across Four Consumer Wearables: Concordance, Temporal Diagnostics, and Signal Harmonization
- Clinical access and scope
- contact our team
Important: This article may discuss BioTwin research, medical vision, regulated clinical pathways, or TwinMe wellness education. TwinMe wellness outputs are not medical or laboratory tests. BioTwin clinical outputs are available only where authorized and through licensed healthcare professionals.