Description

This course covers selected statistical inference and machine learning algorithms for computational genomics. Why join the course? Machine learning is an exciting topic, with many major developments in recent years (even just the last year). As you might guess, the advances in NLP (e.g. LLMs) and computer vision (e.g. CNNs) are being leveraged to analyze molecular sequence data. However, if you go read a scientific paper, or even a blog post (check out this recent blog from Dr. Serafim Batzoglou), you will find that much of the text focuses on concepts from computational biology and uses domain specific terminology like alignment, phylogeny, coalescence, selection, coevolution, etc.

CMSC 498Y is for undergraduate computer science students who are interested in the intersection of machine learning and computational biology -- and who want to get hands-on experience working with models and data. The course material will cover classical machine learning algorithms for molecular sequence data (many of which are based on hidden markov models) as well as more recent developments (based on LLMs and CNNs), along with key background from computational biology. The course will be divided into modules focusing on specific prediction tasks. For each task, we will consider (1) the biology relevant to the prediction task, (2) the input data as well as how it is generated and curated for supervised training, and (3) the models used for prediction and how they are evaluated.

This year, there will be four modules:

1. Basic models of biological sequences
2. Protein family prediction
3. RNA secondary structure prediction
4. Protein secondary and tertiary structure prediction

CMSC 498Y counts towards the CS Undergraduate Major requirements under Area (2) Information Processing.

Web Accessibility

Please provide feedback on web accessibility to the instructor.