6.S084, 6.884, 20.S940, 20.S948 Computational Aspects of Therapeutic Design

9/26/2019 Announcement: PROBLEM SET 1 is posted

Problem set 1 is now posted. The code with a README is available on Github.

9/26/2019 Announcement: OCTOBER Paper Signups

The second round of Paper Presentation Slots are now open. If you have not yet presented, please sign up on Google Forms, here.

9/10/2019 Announcement: Paper Presentation Sign Ups

6.884 students: our the first round of sign ups for paper presentations covering September presentation slots are now up. View it here!

Additional course websites:

• MIT Stellar
• Piazza (discussion forum)
• Learning Modules ((where you download problem set assignments and upload your solutions)

Course description

Welcome to an exploration of computational challenges in the design of human therapeutics. We will explore computational methods in the design and analysis of therapeutics, including small molecules, biologics, cell based therapies, and synthetic biology approaches. Lectures will present essential computational methods on molecular design and optimization drawing upon recent results in machine learning. Classes will include presentations by students on recent research results related to the computational design of therapeutics and efficacy. Problem sets will explore computational methods for therapeutic design. Topics include protein design, antibody optimization, small molecule design and characterization, and the engineering of viruses and cell lines for therapeutic effect. Experts from industry will present on their views of the promise of computational approaches, what is working, and what is needed.

As part of the subject students taking the graduate version will use cloud resources to implement solutions to problems in therapeutic design, first in problem sets that span a carefully chosen set of tasks, and then in an independent project. You will be programming using Python 3 and TensorFlow 1.12 in Jupyter Notebooks on the Google Cloud, a nod to the importance of carefully documenting your work so it can be precisely reproduced by others.

Syllabus and schedule

Date	Type	Title	Speaker	Role	Affiliation	Presentation Date	Papers
9/5	Invited Speaker	Disease Phenotype Identification	Holger Hoefling	Lead of Scientific Data Analysis - Machine Learning and Quantative Analysis	Novartis		Clinical-grade computational pathology using weakly supervised deep learning on whole slide images Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer A survey on deep learning in medical image analysis
9/10	Lecture	Overview of Target Identification	David Gifford	Professor	MIT
9/12	Invited Speaker	Systems Biology Based Target Identification	Ernest Frankel	Professor	MIT		Proteomics, Post-translational Modifications, and Integrative Analyses Reveal Molecular Heterogeneity within Medulloblastoma Subgroups Revealing disease-associated pathways by network integration of untargeted metabolomics The transcriptomic landscape and directed chemical interrogation of MLL-rearranged acute myeloid leukemias
9/17	Presentation	Presentation on Computer-aided Drug Design (9/26 prep)
9/19	Presentation	Presentation on Small Molecule Design (10/8 prep)
9/24	Presentation	Presentation on Small Molecule Design (10/10 prep)
9/26	Invited Speaker	Using computational methods to address druggability and drug discovery challenges	Jose Duca	Global Head of Computer-Aided Drug Discovery	Novartis	9/17	Allosteric inhibition of SHP2 phosphatase inhibits cancers driven by receptor tyrosine kinases Using Membrane Partitioning Simulations To Predict Permeability of Forty-Nine Drug-Like Molecules Small-molecule WNK inhibition regulates cardiovascular and renal function
10/1	Presentation	Presentation (10/3 prep)
10/3	Invited Speaker	Introduction to drug development and the role of quantitative sciences”	Birgit Schoeberl	Global Head Modeling and Simulation, PK Sciences	Novartis Institutes for BioMedical Research	10/1	Clinical responses to ERK inhibition in BRAFV600E-mutant colorectal cancer predicted using a computational model. Mathematical models of tumor cell proliferation: A review of the literature. A Translational Quantitative Systems Pharmacology Model for CD3 Bispecific Molecules: Application to Quantify T Cell-Mediated Tumor Cell Killing by P-Cadherin LP DART Predicting ligand-dependent tumors from multi-dimensional signaling features
10/8	Invited Speaker	Junction Tree representations of small molecules	Tommi Jaakkola			9/19	Junction Tree Variational Autoencoder for Molecular Graph Generation
10/10	Invited Speaker	ML-driven small molecule selection in drug discovery	Jeremy Jenkins	Head of Data Science in Chemical Biology & Therapeutics,	Novartis	9/24	Using information from historical high-throughput screens to predict active compounds. Enrichment of high-throughput screening data with increasing levels of noise using support vector machines, recursive partitioning, and laplacian-modified naive bayesian classifiers. Integrating high-content screening and ligand-target prediction to identify mechanism of action.
10/15	Holiday	Enjoy The Holiday!
10/17	Invited Speaker		Alex Zhavoronkov,	Founder and CEO	Insilico Medicine	9/19	Deep learning enables rapid identification of potent DDR1 kinase inhibitors Molecular Sets, MOSES: A Benchmarking Platform for Molecular Generation Models In silico Pathway Activation Network Decomposition Analysis (iPANDA) as a method for biomarker development
10/22	Presentation	Syn Bio presentation (10/24 prep)
10/24	Invited Speaker	Synthetic Biology Approaches to Disease Therapeutics	Ron Weiss	Professor	MIT	10/22	Genetically engineering self-organization of human pluripotent stem cells into a liver bud-like tissue using Gata6 Programming gene and engineered-cell therapies with synthetic biology Multi-Input RNAi-Based Logic Circuit for Identification of Specific Cancer Cells
10/29	Presentation	Presentation on Vaccine Design (10/31 prep)
10/31	Invited Speaker	Design of Peptide Vaccines	Cathy Wu	Professor of Medicine	Dana-Farber Cancer Institute and Harvard Medical School	10/29	An Immunogenic Personal Neoantigen Vaccine for Patients with Melanoma Mass Spectrometry Profiling of HLA-Associated Peptidomes in Mono-allelic Cells Enables More Accurate Epitope Prediction. Towards personalized, tumour-specific, therapeutic vaccines for cancer.
11/5	Presentation	Presentation on cell therapy (11/7 prep)
11/7	Invited Speaker	Regenerative cell based therapies	Doug Melton	Professor	Harvard	11/5
11/12	Lecture	Machine Learning-based Antibody Design	David Gifford	Professor	MIT
11/14	Invited Speaker	CRISPR Therapeutic Strategies`	Han Altae-Tran	Graduate Student	Broad Institute
11/19	Invited Speaker	Cellular immune therapies	Michael Birnbaum	Professor	MIT	11/21
11/21	Presentation	Presentation on immune therapies (11/19)
11/26	No class
11/28	Holiday	Enjoy the Holiday!
12/3	Project Presentations
12/5	Project Presentations
12/10	Project Presentations

Prerequisites

Undergraduate version: Fundamental knowledge of machine learning, programming, and biology (GIR level). You should be comfortable with calculus, linear algebra, (Python) programming, probability, and introductory molecular biology. Graduate version: Understanding of machine learning with Python and commonly-used libraries. The graduate version is targeted towards students with a high degree of fluency in Computation and Biology and to fully understand the material, students will be best off having previously taken machine learning or computational systems biolog (6.874) or a similar course.

Class meeting times and places

• Lecture: TR 11AM-12:30PM MIT Room 56-114
• Lecture: T 2PM-3PM MIT Room 26-328

Contact

The best way to get detailed questions answered is to visit TA office hours or post them on Piazza.

Office hours

Grading

Class presentations (30%), programming-intensive problem sets (30%), and a final project (40%). Attendance in lecture is important as the class moves quickly and you will need to be present. There will be a requirment for students to give presentations on relevant research papers from the syllabus list. If you will need special accomodations, please email Student Support Services - S3). We will be happy to accomodate!

Project

Final project details TBA

Papers

Lectures will be given by MIT professors and guests from industry on current topics at the forefront of research in computational therapeutics design. These lectures will be accompanied by academic papers listed in the course schedule and available through the Stellar site