Spring 2022 -- CS 590-02/CBB 590-01 -- Tuesday 1:45 Pm -- Prof. Bruce Donald
Abstract: This year's seminar will focus on COVID-19, SARS-CoV-2, and Machine Learning Approaches to Therapeutic Design. However, background material and fundamental basic science, algorithms, and computational methods will be covered in order to get to the foothills of this field.
Some of the most challenging and influential opportunities for Physical Geometric Algorithms (PGA) arise in developing and applying information technology to understand the molecular machinery of the cell. Recent work shows that PGA techniques may be fruitfully applied to the challenges of structural molecular biology and rational drug design. Concomitantly, a wealth of interesting computational problems arise in proposed methods for discovering new pharmaceuticals.
Recently, the intriguing success of Alpha Fold 2 from DeepMind of Google Alphabet has energized the field to pursue machine learning techniques for protein structure prediction and molecular design. Among other topics, we will read some papers on these techniques from the intersection of machine learning and computational structural biology.
This seminar course focuses on topics in computational biology. We will emphasize themes that unite algorithms, modelling, machine learning, or experimental results. Topics will include algorithms, modeling, machine learning, and experimental validation for several areas, including protein design, protein:protein interactions, structural biology, structural immunology, and structure-based drug design.
For those who have taken a class or seminar with me previously, this semester we will read entirely different papers, so please feel free to sign up.
Click here for class overview, schedule, details, etc.
Recitation Time & TA Office hours: Tuesday 6 PM and Thursday 1:30
PM
Location: Languages 109
We will meet:
First meeting: Tuesday, January 11.*
*However, by order of Duke Provost, our class on 1/11 will meet via Zoom. Join: https://duke.zoom.us/j/96689927635. This requires NetID/Shibboleth authentication to join: i.e., logged in as a Duke user.
Class Webpage: www.cs.duke.edu/donaldlab/Teaching/Seminar22/
Some of the most challenging and influential opportunities for
Physical Geometric Algorithms (PGA) arise in developing and applying
information technology to understand the molecular machinery of the
cell. Recent work shows that PGA techniques may be fruitfully applied
to the challenges of structural molecular biology and rational drug
design. Concomitantly, a wealth of interesting computational problems
arise in proposed methods for discovering new pharmaceuticals.
Recently, the intriguing success of Alpha Fold 2 from
DeepMind of Google Alphabet has energized the field to pursue
machine learning techniques for protein structure prediction and
molecular design. Among other topics, we will read some papers
on these techniques from the intersection of machine learning and
computational structural biology.
This seminar course focuses on topics in computational biology. We
will emphasize themes that unite algorithms, modelling, machine
learning, or
experimental results. Topics will include algorithms, modeling,
machine learning, and
experimental validation for several areas, including protein design,
protein:protein interactions, structural biology, structural
immunology, and structure-based drug design.
NB: This course is cross listed as Compsci 590 and CBB 590.
For those who have taken a class or seminar with me previously, this semester we will read entirely different papers, so please feel free to sign up.
Graduate students and undergraduate students are welcome in this class. In this class I welcome students from diverse backgrounds: computer science, biochemistry, biology, chemistry, engineering, physics... It is recommended that students be interested in the connections between computational science and the life sciences as applied to macromolecules of biological and pharmacological importance.
In this seminar course students will present both recent and classic papers from the literature, and also compile notes on these papers.
The primary reading for this course will be supplied as papers to the students. While some of the background for these papers may be unfamiliar, the class is structured so that students can acquire this background while preparing to present and discuss the papers. Specifically, students will read a textbook, that is designed for this course, in order to prepare for and understand the background to present the papers. One textbook covers basic algorithms in this area of computational biology, and their applications. The second covers recent results in the field of protein design. When the weekly papers are assigned, relevant chapters of the textbooks will be assigned as area/background reading. However, student presentations will concentrate on the papers, not on presenting from the textbooks.
To give you an idea about the kind of papers we will read, here is the
schedule from last year. Note that we will read different papers
this year! This is just to give you an idea!
Textbooks:
Overview
| Syllabus
| Schedule
| How to give a good talk
Supplemental Materials
| Some Relevant WWW Links
| Recitation Materials
Course Summary/Syllabus:
"Strictly speaking, molecular biology is not a new
discipline, but rather a new way of looking at organisms
as reservoirs and transmitters of information. This new
vision opened up possibilities of action and intervention
that were revealed during the growth of genetic
engineering."
- Michel Morange,
"A History of Molecular Biology," Harvard University Press.
This year's seminar will focus on COVID-19, SARS-CoV-2, and
Machine Learning Approaches to Therapeutic Design.
However, background material and fundamental basic science,
algorithms, and computational methods will be covered in order to get
to the foothills of this field.
