MADRID--(BUSINESS WIRE)--Oct 23, 2026--
Genialis, the therapeutic intelligence company, will present data at the European Society for Medical Oncology (ESMO) Congress 2026 showing that its biology-first models predict treatment response even with limited available patient data. Across three distinct cancer types and treatment regimens, the models demonstrated superior predictive performance in cohorts of fewer than 70 patients, addressing one of the fundamental challenges in developing predictive biomarkers for precision oncology.
This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20261022380424/en/
Clinical trials are small data. Early-stage studies often enroll relatively small patient populations, creating challenges for conventional modeling approaches attempting to extract meaningful predictive signals. Yet drug developers must make critical decisions about pipeline advancement and patient selection criteria, often before independent validation data are available.
A tumor RNA-seq profile measures roughly 20,000 genes; with 50 patients, a conventional deep learning model risks identifying statistical noise rather than meaningful biological signals. Overcoming this “curse of dimensionality” has been central to Genialis’ approach to predictive modeling since its founding.
The Genialis™ Supermodel is a large molecular model of cancer biology trained on hundreds of thousands of transcriptomic samples. The Supermodel transforms tens of thousands of individual gene-expression measurements from RNA sequencing (RNA-seq) into biologically meaningful features called “biomodules,” each representing a fundamental aspect of cancer biology, including oncogenic signaling, immune activation, and DNA damage response. By reducing the complexity of molecular data into interpretable biological features before model training, Genialis can develop predictive biomarkers from relatively small patient populations while providing insight into the biological mechanisms driving treatment sensitivity and resistance.
The ESMO poster (abstract #8664) evaluates this approach across datasets from three different cancer types in a typical size of early clinical studies: pretreatment RNA-seq from anti-PD-1-treated melanoma (n=41), pembrolizumab-treated gastric cancer (n=45) and KRAS G12C-mutant non-small cell lung cancer (NSCLC) treated with sotorasib (n=66). Under strict held-out testing, biomodule models outperformed matched gene-expression models in every cohort. In NSCLC, biomodules predicted sotorasib response with an AUROC of 0.8, against 0.6 for the best gene-expression model, and predicted responders stayed on treatment a median of 337 days versus 178. Each prediction came with a biological rationale, with predicted NSCLC responders showing MAPK-addicted biology, while non-responders exhibited PI3K bypass programs.
From Small Cohorts to an N-of-1 Application: The Sid Sijbrandij Case Study
When GitLab co-founder Sid Sijbrandij was diagnosed with osteosarcoma in November 2022, he faced a rare bone cancer diagnosed in roughly 1,000 people each year in the United States. When his cancer returned and standard treatment options ran out, Sijbrandij built a team to pursue an extraordinary approach to understanding and treating his disease. Combining what they call “maximal diagnostics” with personalized experimental therapies, the team collected and publicly shared more than 30 terabytes of Sid’s clinical and molecular data at osteosarc.com.
Genialis retrospectively analyzed four tumor RNA-seq libraries generated over 25 months by three clinical laboratories from different specimens. The Supermodel enabled consistent biological comparisons across the four samples, even where raw gene-expression data were not comparable. By linking each sample's biology to therapeutic hypotheses, the analysis demonstrated how researchers can revisit those hypotheses as additional samples become available. Genialis expects this approach to become increasingly valuable as precision oncology advances toward smaller, more precisely defined patient populations and, ultimately, treatment strategies tailored to individual patients.
“For patients facing rare cancers and researchers working with limited patient data, the ability to compare tumor biology across samples collected over time opens new opportunities to investigate disease and develop therapeutic hypotheses,” said Rafael Rosengarten, PhD, Chief Executive Officer and Co-Founder of Genialis. “Together, the ESMO findings and Sid’s case study demonstrate the potential of biology-first AI to extract meaningful insights from limited patient data, whether across a clinical cohort or within a single patient's cancer journey.”
Poster details
Title: Biology-first modeling enables interpretable patient-selection strategies from small oncology cohorts
Abstract: #8664
Date and time: Poster Display session, Sunday, October 25 | 12:00–12:45 CET | ESMO Congress 2026, Madrid, October 23–27, 2026
Poster Category: AI for clinical research and drug development
Presentation no.: 152P
Single-patient analysis:https://www.genialis.com/2026/10/19/a-biology-first-approach-to-precision-medicine-tracking-sids-tumor-biology-with-the-genialis-supermodel/
About Genialis
Genialis provides therapeutic intelligence with its Supermodel of cancer biology. We develop and validate clinically actionable AI models informed by the world’s most ethnographically diverse cancer data sets to predict patient responses and guide treatment decisions for targeted inhibitors, immunotherapies, and other emerging therapeutic classes. Genialis is trusted by pharma and diagnostics partners, and together, we are transforming medicine through data. The company is headquartered in Houston, Boston and Ljubljana.
ESMO 2026 Poster
