BIOCHEM 9552: Quantitative Methods in Biology
Location and Times
Wednesdays 11:30am-1pm
MSB 282
Course Description
Quantitative approaches are now ubiquitous used throughout biomedical research and it has become
increasingly important for biomedical researchers to include training in advanced statistics, computation
and data science in order to maintain modern research programs. This course is designed to provide
biochemistry graduate students with the opportunity to efficiently learn quantitative tools and techniques
relevant to their research regardless of their prior training background. Topics include programming,
software development, permutation and randomization testing, stochastic modelling, machine learning,
artificial intelligence, generative modelling, data cleaning, and visualization.
Learning Outcomes
At the completion of the course, students will:
- have the capacity to independently and efficiently learn new or more advanced computational
tools and languages;
- understand the fundamental principles of data science in the biomedical sciences and will have
developed good practice and procedures for handling data of all types;
- be fluent in the main paradigms of machine learning, deep learning and artificial intelligence with
practical skills to develop learning datasets, to rationally choose appropriate methodology from
these fields that suite the structure and size of their available data; and
- have confidence to soundly address statistical and computational challenges in the context of
their research including the capacity to read the primary literature from these areas.
Course Structure
Students who earn 100 points worth of training exercises receive full credit for the course (0.5).
Course Point Stucture
The course will be run across both the Fall and Winter semester. The course consists of a series of modules covering different
topics and of varying length. For example, a short module lasting a single week could cover one specific
software system that is commonly used in programming and software design (e.g. GIT). Mid-length
modules, which could last 2-5 weeks, cover topics such as a specific computational paradigm (e.g. the
Expectation-Maximization algorithm) or statistical technique (e.g. Frequentist’s confidence interval versus
Bayesian’s credible interval). Full-length modules cover broader topics such as introducing a programming
language, specialized packages (e.g. PyTorch, Stan), working together through a statistics textbook (e.g.
Modern Statistics for Modern Biology (2019) Homes and Huber), or an in-depth coverage of specific topic
(e.g. paradigms in deep learning).
Some modules may be offered outside of standard lecture times and in a compressed time frame (e.g. a
full day workshop).
In the future, we will encourage other quantitative, computational biology and data
science groups to contribute modules. For example, material that might have only been taught within a
group’s research lab meeting might now be offered as a module for credit; these added offerings have
clear benefits for Biochemistry graduate students. Alternatively, invited lecturers and seminars from
outside of the Department of Biochemistry that cover keep topics with value to Biochemistry graduate
students could be developed into a (small) module by including a short grading exercise. Modules could
be created dynamically in response to rapid new developments (certainly an issue given the current rate
of developments in AI) or in response to needs voiced by specific students or groups.
Students will be graded at the end of each module; each module will be assessed a certain number of
points depending on its duration, difficulty and number of hours required for assignments or other
exercises; a pass awards the student this number of points; the student will need to earn sufficient
points (e.g. 100) to complete the course.
For example, a student might choose to complete a full-length module on PyTorch worth 40 points, two
mid-length modules (e.g. Data Science in Python and Probabilistic modelling) lasting 5 weeks each and
worth 20 points each, and one short module that covers the basics of working with the *nix command
line.
(C) M. Hallett, Western University, 2024