Machine learning and AI are increasingly important tools for geneticists, including those working in clinical settings, who need to understand the possible impact of a new or rare genetic variant on patients and disease. But how well do these models actually work? Are they always reliable? Do they have blind spots? In this project, you will explore the performance of state-of-the-art machine used to predict the effects of genetic variation on humans. You will compare their predictions with those from laboratory experiments and other approaches used in genetics. You will gain the ability to critically assess their strengths and limitations, and become familiar with how these models—and the large datasets they rely on—are changing how we understand health and disease, and may possibly transform healthcare in the future.
You don’t need any background in machine learning or programming to take part in this project—just curiosity and a willingness to learn something new as you go.
Supervised by
Available for Student Supervision