At Biodesix®, we use the Diagnostic Cortex® platform, based on modern machine learning techniques to design tests that are reproducible, robust, and answer critical clinical questions for our partners.
Our Solution to Your Biomarker Discovery Challenges
How often has your biomarker strategy for a drug in your pipeline not validated in clinical trials? Classical biomarker discovery uses a hypothesis-dependent approach based on presumed biological rationale for the drug, but unfortunately, there remains a high risk of clinical failure.
At Biodesix, we are at the forefront of personalized medicine using a hypothesis-independent approach to biomarker discovery. We are able to incorporate multi-omic biomarker data, clinico-pathologic characteristics, and clinical outcome data into our machine learning platform to design tests that identify patients likely to benefit from your therapies. Once a test is developed, a biological pathway correlation is performed to help relate to your drugs mechanism of action.
Biodesix’s Test Design Approach
Biodesix’s proprietary machine learning platform, the Diagnostic Cortex, incorporates patients’ multi-omic data, clinico-pathologic characteristics, and clinical outcome data to develop robust clinical diagnostic tests for drugs in your pipeline.
Test Design Inputs
Test Design: The Diagnostic Cortex®
The Diagnostic Cortex platform incorporates traditional machine learning concepts and advances in deep learning. It was optimized for test development in cases with more attributes than samples, with the goal of minimizing the potential for overfitting and promoting the ability of the developed tests to generalize to unseen datasets.
Clinical Diagnostic Test
Biodesix successfully develops both standalone and companion diagnostic tests for target populations, realizing the potential of precision medicine. Our tests support treatment decisions including the selection of immunotherapies, novel therapy combinations, or alternative treatment pathways.
IMMUNOTHERAPY CASE STUDY
A test developed for anti-PD-1 therapy in 2nd line NSCLC that identifies a group of patients that likely won’t demonstrate long term benefit from nivolumab and is likely predictive for nivolumab vs. docetaxel.
Development (nivolumab) (N=98)
Validation (nivolumab) (N=32)
Evaluation (chemotherapy) (N=68)
Source: Goldberg 2017 (SITC poster)SEE DATA
Biological Pathway Correlates
Biodesix discovers hypothesis-independent tests with machine learning that answer critical clinical questions and identify distinct patient groups. The question then becomes: what is the biology behind the classification of these patient groups? Answering this question reveals the biological underpinnings for the test; it can also inform combination therapies or even suggest expansion indications for patients who have similar underlying disease biology.