Advanced seminar on computational methods in the design and analysis of human therapeutics, including vaccines, small molecules, biologics, cell based therapies, and synthetic biology approaches. Lectures will present recent results on computational methods on vaccine design, molecular design and optimization drawing upon advances in machine learning. Recitations will include presentations by students on recent research results related to the computational design of therapeutics and efficacy. Topics include protein design, antibody optimization, vaccine design, small molecule design and characterization, and the engineering of viruses and cell lines for therapeutic effect. Experts from industry and faculty will present their views of the promise of computational approaches, what is working, and what is needed.
Submit PDF of your PSET1 here by May 3rd.
As announced during the recitation on Tuesday 4/19/2022, there will be no in-person recitation on Tuesday 4/26/2022.
Needless to say, you must still read the papers, fill-out recitation check-in form to submit questions for the guest speaker, and come to the lecture on 4/28/2022, when we will have a fascinating talk by Dr. Matt Hayward.
Problem set 2 is due at 11 AM on Tuesday, May 3rd, 2022. Please follow the following instruction carefully.
(1) Go to the following link.
(2) Click "File" > "Save a copy in Drive" to copy the Google Colab template to your Google drive.
(3) In the coding portion of the exercises, you will find strings "fill_in_the_blank". You need to edit these strings appropriately to make the neural network up and running.
(4) If the neural network is up and running, you will produce two plots, each standing for different regime of training vs. test set splits. In a seperate PDF, you will attach the plots and discuss the consequences of different splitting regimes
(5) As in PSET1, you can work with a partner or by yourself in coding portion. The written portion of the work (discussion in PDF) should be your own work, however.
Your TA will not able to hold his usual office hour on Monday, 4/11, at 3pm. Please email him (hyunjinp@mit.edu) if you want to schedule for an alternative OH.
We understand that everyone's schedule is getting hectic as the semester marches on, but recitation attendance is getting underwhelming. We encourage everyone to come to recitations for three reasons. First, recitation is an integral component of learning experience of the course, as it encourages critical and active reading of scientific publications. Second, it is a courtesy due to your fellow student presenters, as it feels quite dismaying to present to a sparsely populated room. Third, you will find out your PSET grades iff you're at the recitation. So, come join us on Tuesdays at 11!
Submit PDF of your PSET1 here by March 15th.
Per MIT policy, February 22nd, 2022 will be held as Monday's class. We therefore shall not have in-person recitation as usual. Still, please fill in mock check-in before 2/22 with questions to ask to Dr. Andrian Veres.
Problem set 1 is due at 11 AM on Marth 15th, 2022. Download pfizer and moderna vaccine sequences. Notice that all 1-methyl-3'-pseudouridylyl bases are replaced by thymidines. You can do it yourself or form a 2 or 3-people team. If you decide to form a team, please list all your collaborators. P.S. If you are dying to know more about mRNA vacines against SARS-CoV-2 viruses, you can read New Engliand Journal of Medicine's articles on Pfizer and Moderna vaccines and/or FDA Letters of Authorization for Pfizer and Moderna vaccines. Doing so is absolutely not necessary.
Welcome to Machine Learning Based Therapeutics Design! Please complete the first-day survey, whether you're taking the course for credit, registered as a listener, or simply here for fun. If you're taking the course for credit, please sign up for paper presentation.
| Date | Title | Speaker | Role | Affiliation | Papers | WebCast Link | Recitation |
|---|---|---|---|---|---|---|---|
| 02/03 | A new paradigm for vaccine evaluation, design, and its application to SARS-CoV-2 | David Gifford | Professor | MIT | Liu2020a Liu2021 Carter2021 | Link | No recitation |
| 02/10 | Synthetic, biodegradable lipid nanoparticles for gene therapy and gene editing | Qiaobing Xu | Professor | Tufts University | Qiu2021 Chen2021 Ma2020 | Recording | check-in (2/8) presentation |
| 02/17 | COVID-19: Developing a Vaccine During a Pandemic | Dan Barouch | Professor | Harvard Medical School, Center for Virology and Vaccine Research | Liu2022 Gebre2021 | Recording Presentation | check-in (2/15) presentation |
| 02/24 | Regenerative medicine for Type 1 Diabetes | Adrian Veres | Postdoctoral Fellow | Harvard University | Veres2019 Yeo2021 Qiu2022 | Recording Presentation | mock check-in |
| 03/03 | Data-driven small molecule design and synthesis | Connor Coley | Professor | MIT | Coley2018 Graff2021 Gao2021 | Recording Presentation | check-in (3/1) |
| 03/10 | Using computational methods to address druggability and drug discovery challenges | Jose Duca | Global Head of Computer-Aided Drug Discovery | Novartis | Chen2016 Dickson2018 Yamada2016 Schneider2019_optional | In-person only | check-in (3/8) |
| 03/17 | ML-driven small molecule selection in drug discovery | Jeremy Jenkins | US Head, Chemical Biology & Therapeutics | Novartis | Schneider2017 Godinez2021 Schneider2019_optional | In-person only | check-in (3/15) |
| 03/24 | Spring Break | Enjoy your break! | |||||
| 03/31 | Synthetic Biology Approaches to Disease Therapeutics | Ron Weiss | Professor | MIT | Siciliano2018 Xie2011 Li2020 | Recording | check-in (3/29) |
| 04/07 | Antibody CDR Design Using High-Capacity Machine Learning | Stefan Ewert | Associate Director | Novartis | Liu2020b Dai2021 | Link-no recording will be posted per Novartis policy, so please attend real-time! | check-in (4/5) |
| 04/14 | Antigen recognition in immunotherapies: TCR and signaling domain engineering | Michael Birnbaum | Professor | MIT | Kyung2021 Dobson2021 Joglekar2021 | Recording | check-in (4/12) |
| 04/21 | Using deep learning to decode the genome and 'program' life-changing medicines | Brendan Frey | Founder & CEO | Deep Genomics | Merico2020 Xiong2015 Alipanahi2015 Wainberg2018 | Link | check-in (4/19) |
| 04/28 | Vaginal microbiome, mucosal immunology, and HIV in sub-Saharan Africa | Matt Hayward | Postdoctoral fellow | Harvard Medical School, the Kwon lab | Gosmann2017 Munoz2021 Bloom2021 | Link | check-in (4/26) |
| 05/05 | CRISPR and protein design | Han Altae-Tran | PhD student | Broad Institute of MIT and Harvard | Altae-Tran2021 Saito2021 Shen2018_optional | Link | check-in (5/3) |
There will be two problem sets. Problem Set 1 explores the design of contemporary mRNA COVID-19 vaccines. Problem Set 2 explores the use of deep learning for antibody design. Students will also present papers during recitation. If you will need special accomodations, please email Student Support Services: s3-support@mit.edu. We will be happy to accomodate!