Client success story
Multi-Modal Machine Learning For Target Discovery
Business Value Delivered: Multiple Actionable Drug Targets
Summary: Artificial Intelligence-Enabled Multi-modal Target Discovery
bPrescient built a scalable AI framework that initially integrates imaging, molecular and clinical data sets, allowing for identification of precision phenotypes and novel therapeutic targets.
Challenge: Identifying Robust Targets In Complex and Noisy Data
An early-stage therapeutics company in the neuroscience space is identifying novel drug targets and is faced with a complex set of noisy data from multiple modalities, including imaging, molecular and clinical. The company needs a scalable framework for running AI and machine learning analytics that can handle missing data, noisy data and multiple data sources and modalities. Inability to handle these complexities would lead to spurious results and waste research and development time in pursuing them.
Solution: AI And Machine Learning Identified Patient Subgroups Associated With Targets
bPrescient’s team of AI subject matter experts and machine learning model developers worked with the company stakeholders to design and build a cloud platform that incorporates imaging, molecular, and clinical data into knowledge-driven and hypothesis-free machine learning methods with the goal of identifying actionable patient clusters that can be the source of novel targets and biomarkers for patient type and status. This powerful framework was designed to (a.) incorporate new data modalities without retraining and (b.) robustly handle missing data, both of which are often a challenge in patient data sets.
Outcome: Identification Of Targets And Biomarkers Associated With Clinical Subgroups
The company used the scalable framework to analyze multiple data sets and identify patient subgroups and molecular markers that are being explored as clinical trial inclusion criteria and potential therapeutic targets. The results were instrumental in achieving a next round of funding to develop those therapeutics against the novel targets.
Let’s talk
Tell us about your toughest R&D challenge and let’s discuss how bPrescient can solve it.
