We transform transcriptomic data into actionable patient subgroups and treatment recommendations

We model fundamental biology using human data and machine learning to ensure successful development of new drugs, and inform targeted treatment decisions for better outcomes.

We are building the next generation of cancer biomarkers based on gene expression data. They incorporate dozens to hundreds of genes into machine learning classifiers capable of capturing therapeutically meaningful relationships between genes and pathways.

Download the whitepaper

Answer questions like:

How do our preclinical findings translate to patients?
What’s this drug’s mechanism of action?
How do different patient populations respond biologically?
What drives response or resistance?
How durable is the response?

Genialis algorithms clearly discern responders from non-responders on a biomarker probability plot based on RNA biomarkers.

RNA Biomarkers:
The Future is Now

Why RNA?

Genialis focuses on RNA because gene expression-based biomarkers give us the best of both worlds: insight into disease drivers at the genome level and dynamic states defined by the transcriptome. The additional information on pathways and processes, especially compared with DNA or protein biomarkers, are especially relevant to predicting treatment outcomes of complex diseases.

Genialis prefers RNA biomarkers for several reasons

RNA is closer to the phenotype than DNA. Assessment of RNA reveals changes in gene expression associated with disease progression, mutation, drug response, chemical perturbation, etc
RNA can be used to infer mutation status of coding genes in the same manner that DNA can.
Data from total RNA sequencing contains high-dimensional information which can be leveraged to understand the biology of a given cancer with far more nuance than DNA alone.
The standardization of RNA sequencing, the ability to multiplex many different tests from the same analyte combined with the ever-decreasing costs of sequencing makes transcriptomic biomarkers cost-effective for complex disease diagnosis, and clinical R&D.

How biomarkers predict response to targeted treatment

	DNA biomarker	RNA biomarker
Biology	Driver gene disruption, mutational burden	Pathway disruption, dysregulation, activation, suppression, etc PLUS: Driver gene disruption, mutational burden
Feature	One or just a few DNA variants	Quantitative signatures comprising the expression and variants of dozens of genes
Algorithm	Binary mutation status	Machine learning classifier
Capability	Necessary but insufficient: Describes the status of a drug target or disease driver, but does not describe the biological state (or phenotype) of the disease.	Predictive and explanatory: Captures complex interactions to define the biological state and treatment susceptibility of the disease.