Skip to content

Characterizing disease by molecular mechanism

Chronic diseases present the greatest challenge to human health. They're caused by a complex interplay of biology, which gold standard genomic analysis and AI models are unable to detect. 

Hypothesis-free combinatorial analytics is uniquely able to reveal how multiple genes and other factors interact to cause complex diseases. Identifying the mechanisms specific to each patient to personalize care and improve health outcomes.

Our platform is unbiased and infinitely scalable, producing findings that are reproducible and explainable – making new discoveries faster than ever, with the highest level of confidence.

Combinatorial analytics vs GWAS-min

Combinatorial disease signature comprising six co-associated SNPs

2.ALZHEIMERS-min

Going far beyond all other data models 

LLMs and other AI models are limited by their training sets, which are biased and too incomplete to address chronic disease biology.

Combinatorial analytics enables us to rapidly analyze multiomic datasets at massively higher orders of combinations – far more than is possible with even the most advanced data models.

We capture non-linear signal in the biology of genetic and metabolic networks – revealing the effects of combinations of features and the deepest insights into disease biology to de-risk every stage of drug development and clinical care.

Going beyond AI-min

Analyzing every combination to find the cause

Our platform rapidly analyzes the widest range of multimodal data types, singly or in combination, without bias to identify features that in combination are associated with a phenotype.

This allows us to generate much more and far richer signal from smaller datasets than traditional Genome Wide Association Studies (GWAS) models and to better stratify patients according to the mechanistic origins of their disease.

 Multiomic
 Genomic
 Metabolomic
 Proteomic
 Transcriptomic
 Phenotypic 
 Microbiome
 Epidemiological 

 Clinical data
 Imaging 
 Environmental 
 Pathogens 
 Structural variants 
 Digital biomarkers 
 Real world evidence
 more

Deeper disease understanding & higher replication

Combinatorial analytics identifies significantly more mechanistic insights, from less data, with considerably greater replication than any other analytical method.

The graphic below shows this in a direct comparison between the PrecisionLife platform and a meta-GWAS of endometriosis patient data, published in Nature Genetics¹ by an international consortium of key opinion leaders.

Combinatorial analytics disease insights
CASE STUDY: COVID-19 

More disease insights, faster

Our platform outperformed global efforts to identify the genetic factors of severe COVID-19 susceptibility, producing more insights, clinical biomarkers, and novel drug targets than any other analysis before or since.

Completed 12-months earlier than the COVID-19 Host Genetics Initiative (HGI), our analysis identified many more loci and more than 7X the number of novel genes associated with severe COVID-19 – compared to any other global research (as shown in the figure below) – and linked them all to major symptoms of the disease.

Our findings have since been extensively validated independently by research groups around the world.

Read the first publication of these results in June 2020

Updated-PL-COVID-19-comparison-table-1
PAPER

Combinatorial analytics: an essential tool for the delivery of precision medicine

Artificial Intelligence in the Life Sciences

 

Read the paper

Patient Stratification Biomarkers

Our mechanistic patient stratification biomarkers guide drug developers and healthcare professionals towards the most appropriate treatment strategies for individual patients.

Our biomarkers

DiseaseBank™

DiseaseBank is the richest repository of mechanistic insights generating chronic disease IP for novel targets, biomarkers, clinical trials design/rescue and diagnostics.

Our DiseaseBank