Context

Genome sequencing is currently generating a huge amount of data. Issues arise when storing these data and when computing with such data. Analysis pipelines can take days, specially when dealing not only with a person, but with a population (500’000 people have already been sequenced in the UK).

The objective of this project is to develop systems able to take advantage of heterogeneous hardware to perform certain genomic and genetic algorithms. Devices such as FPGAs enable to parallelize computation to get results faster, but also to reduce power consumption. After having worked on some algorithms (pair-HMM, clustering, k-mer counting, PBWT), a new format to store genetic data has been proposed. In a third phase, we plan to implement some processing on a new generic computing platform.

Partners

Prof. Olivier Delaneau, Prof. Ioannis Xenarios, UNIL

HEIG-VD Team

Rick Wertenbroek (PhD candidate), Prof. Yann Thoma