Joseph
Brown
Canada
Area of Research
We study how early “hit” compounds can be matured into potent, stable, and selective drug candidates. Our focus is on molecules that lie between small molecule drugs and proteins, specifically peptides and peptidomimetics, and on improving their key properties: activity, affinity, permeability, and specificity. By linking chemical synthesis with biological testing and machine learning, we aim to transform how new medicines are discovered.
Primary Research Challenge
Modern drug discovery and development faces an constant overwhelming chemical space: a near infinite number of possible molecules to evaluate for every useful one brought to the clinic. We ask how this vast space can be navigated efficiently to find compounds with the right balance of activity and properties for therapeutic use. How can we leverage advances in algorithms, automation, and data to guide and accelerate this path of discovery and optimization?
Proposed Solution
Our solution is a data-driven, AI-guided discovery platform. We are focused on the concept of a self-driving lab that couples robotic synthesis, biological assays, and active machine learning algorithms. We specifically use direct-to-biology methods, such as affinity-selection mass spectrometry (AS-MS), allow us to test molecules without purification and feed immediate results back to machine-learning models that design the next experiments.
Impact to Date
Our lab is new, but to date, we have led and made contributions toward the automation of key biological assays, creation of chemistry for large cyclic-peptide libraries, and constructed bioinformatic tools to increase data throughput. We have applied active-learning strategies to molecular discovery, demonstrating faster, smarter optimization cycles that shorten the path from initial hit to drug-like molecule.
Publications
Notable Awards
- PhRMA Foundation Postdoctoral Fellowship
- NSF Graduate Research Fellowship
Keywords: Drug discovery; Chemical biology; Machine learning; Automation; Self-driving lab; Peptides; Molecular design
