Artificial intelligence is having a broad impact on society, and there is a push to harness these technologies for the purposes of drug discovery and development. This graduate course is a pilot course whose objective is take students through an exploration of the opportunities and challenges in AI for drug discovery and to provide a launchpad for genuinely inter-disciplinary, collaborative projects. It is delivered jointly through two sister courses: CSC2541 in Department of Computer Science and (PCL3107,PCL3108) in the Department of Pharmacology & Toxicology.
The scope of the course includes structural biology, phenomics, and genomics. With facilitation, students will conduct hands-on active learning assignments using simplified datasets to solidify their understanding of drug discovery data and problems. A course project, carried out in teams comprised of students from both departments, is the centrepiece of the course's work. The project will involve co-designing the research questions, research designs, and datasets of authentic projects at the intersection of AI and drug development. Tutorials will involve co-working sessions facilitated with expert guidance.
Students in the Computer Science Department will appreciate the priorities and research approaches of drug discovery scientists. Computer science students will share their domain-specific expertise and their perspectives about modern machine learning algorithms with their counterparts in Pharmacology & Toxicology.
Students in the Pharmacology & Toxicology Department will appreciate the priorities and research approaches of AI scientists. Pharmacology and Toxicology students will share their domain-specific scientific expertise and their perspectives about experimental validation with AI scientists.
CSC Instructor (giving PCL lectures): Chris Maddison
PCL Instructors (giving CSC lectures): Rebecca Laposa and Jean Martin Beaulieu
TAs: Micaela Elisa Consens, Ella Miray Rajaonson, Stefan Vislavski
Email Instructor and TAs: ai4dd@cs.toronto.edu
You can find the syllabus here.
| Assignment | Due | CSC Credit | PCL3107 Credit | PCL3108 Credit |
|---|---|---|---|---|
| Participation | - | 15% | 15% | 15% |
| Module 1 (Boltz) | Jan 23 (due before the tutorial) | 7.5% | 15% | - |
| Module 2 (RxRx3) | Feb 6 (due before the tutorial) | 7.5% | 15% | - |
| Module 3 (Genome LMs) | Feb 20 (due before the tutorial) | 7.5% | 15% | - |
| LOI | Feb 23 | 20% | 40% | - |
| Milestone 1 writeup | Mar 9 | 11.25% | - | 22.5% |
| Milestone 2 writeup | Mar 30 | 11.25% | - | 22.5% |
| Project presentation | Apr 1 to Apr 3 | 20% | - | 40% |
This is a graduate course designed to explore the challenges in opportunities in AI for drug discovery. So, while there are no formal prerequisites, the course does assume a certain level of familiarity with key concepts, depending on the student's home department. Computer science students are expected to be proficient in machine learning, for example through having taken a previous course in machine learning such as CSC311 or STA314 or ECE421. Pharmacology & toxicology students are expected to have familiarity with biostatistics, ideally with large datasets. Students are not required to have prior experience with Machine Learning, but it is helpful to have experience or interest in upskilling in one or more of these areas: Foundations of Linear Algebra, Python, Introduction to Machine Learning.
We are not allowing auditors in this course.
This is a preliminary schedule, and it may change throughout the term.
| Week of | PCL Lectures on Wed (MY440) |
CSC Lectures on Thu (MY360) |
Joint tutorials on Fri (MY380) |
|---|---|---|---|
| Jan 5 | Intro to ML (Maddison) | Intro to DD (Laposa) | Social mixer |
| Jan 12 | Structure Prediction I (Maddison) | Proteins (Laposa) | Pre-module 1 (Boltz) Module 1 (Boltz) Post-Module 1 (Boltz) |
| Jan 19 | Structure Prediction II (Maddison) Phenomics I (Maddison) |
Target-based screening (Laposa) Live cells (Beaulieu) |
LOI assignment and MTA lecture |
| Jan 26 | Phenomics II (Maddison) | Phenotypic screens (Laposa) AI and Genetic I (Beaulieu) |
Module 2 (RxRx3) |
| Feb 2 | Large Genomics Models I (Maddison) | AI and Genetic II (Beaulieu) | LOI co-working |
| Feb 9 | Large Genomics Models II (Maddison) | Emerging therapeutic modalities (Laposa) Gene therapy (Beaulieu) |
Module 3 (Genome LM) |
| Week of | Joint lectures on Wed (Room TBD) |
Joint lectures on Thu (MY360) |
Joint tutorials on Fri (MY380) |
|---|---|---|---|
| Feb 23 | - | LOI feedback and milestone 1 planning | Co-working time |
| Mar 2 | - | Weekly check-in | Co-working time |
| Mar 9 | - | Milestone 2 planning | Material Transfer Agreements |
| Mar 16 | - | Milestone 1 feedback | Co-working time |
| Mar 23 | - | Weekly check-in | Co-working time |
| Mar 30 | - | Project presentations | - |
Week 1
Week 2
Week 3
Week 4
Week 5
Week 6