The Challenge We Set Out to Solve
One of the most persistent challenges in precision oncology today is the interpretation of Variants of Uncertain Significance (VUS):- genetic mutations detected in a patient’s tumour that cannot be immediately classified as harmful or benign.
As comprehensive genomic profiling (CGP) becomes standard practice in oncology, the volume of VUS data is growing rapidly and so is the clinical burden of not knowing what these variants mean for individual patients.
At 1Cell.ai, we believed there had to be a smarter way.
So we built one.
Our Approach
Using our OncoIndx® panel for next-generation sequencing (NGS)-based CGP, we developed a proprietary in-house automated precision classification and interpretation system:- a machine learning model that integrates multiple lines of evidence to systematically reclassify VUS with high clinical utility.
Our system evaluates variants across multiple dimensions simultaneously including; in silico evidence, evolutionary conservation scores, functional RNA studies, patient and family history, and genomic co-findings such as microsatellite instability and protein expression data, to generate a coherent, biologically consistent explanation for disease causation.
What We Found
The results validated the power of our multi-evidence approach. In one illustrative case, several VUS intronic variants, initially flagged as uncertain, were reclassified as likely pathogenic based on in silico evidence (PP5, PM2, BP4) and low evolutionary conservation scores (-0.423). Our system also identified the presence of Lynch Syndrome and associated genomic findings including high microsatellite instability and loss of MSH2 protein on IHC insights that would have been missed by conventional single-database classification approaches.
When complementary lines of evidence are integrated including; functional loss of a variant, segregation-consistent family history, rarity in population databases, and supportive in silico predictions, our system generates a final, high-confidence verdict of reclassification that directly informs clinical decisions.
Why This Matters
VUS reclassification is not a technical exercise, it is a clinical imperative.
For patients, a VUS reclassified as pathogenic can change their entire treatment trajectory, unlocking targeted therapies, identifying hereditary risks for family members, and enabling earlier, more precise interventions. Our automated machine learning system brings speed, consistency, and clinical rigour to a process that has historically been slow, subjective, and incomplete.
This work reinforces our commitment at 1Cell.ai to not just generate data, but to make every data point clinically meaningful.
Presented by Dr. Gowhar Shafi et al | April 21, 2026 | Session PO.PS01.08
Presented at the AACR Annual Meeting 2026, San Diego, CA | April 17–22, 2026 | #AACR26