Efficient algorithms for mining biological datasets on modern Graphics Processing Units (GPUs)
Project Description

Modern GPUs can offer tremendous computing power, which far exceeds CPUs. Thus GPUs may provide solutions for the grand challenge in computational biology which is to efficiently mine massive datasets generated by computers. In this interdisciplinary project, the students will be co-supervised by professors from Chemistry and Computer Science to implement data mining algorithms in GPUs specifically for biological datasets. Gaining insights of these datasets could help understand a wide range of biological processes and many human diseases. Students with sufficient skills in GPU programming are preferred.

Supervisor
SANDER Pedro
Co-Supervisor
HUANG Xuhui
Quota
2
Course type
UROP1000
UROP1100
UROP2100
UROP3100
Applicant's Roles

The students will implement clustering algorithms in GPUs specifically for biological datasets. A couple of examples are K-center and K-means algorithms.

Applicant's Learning Objectives

1. To learn GPU programming for the purpose of general scientific computing
2. To learn how to deal with massive biological datasets generated by computer simulations
3. To gain insights of these biological datasets which help understand many biological processes

Complexity of the project
Moderate