My research focuses on bioinformatics and computational biology, with an emphasis on algorithms for large-scale sequencing data.
Bioinformatics
The main research focus is on Bioinformatics and Computational Biology. My current research projects include developing methods for memory efficient methods for de novo transcriptome assembly and whole genome assembly for mammalian sized genomes. The advent of High Throughput Sequencing (HTS) has enabled us to obtain massive amounts of DNA and RNA data from various *-seq experiments. However working with this data can be nontrivial. In organisms without a known or well annotated reference genome this is requires a considerable computational effort to construct the genome or transcriptome from scratch. One of the major bottlenecks is memory usage that is made worse from the presence of errors in the sequencing reads. My work focuses on developing algorithms and software to reduce the amount of memory needed so that smaller labs can work with novel data without access to a large scale clusters.
GPU based acceleration
Recently we have been working on methods for accelerating processing of
data in genomics and genetics. In particular we have worked on a GPU
based method for pseudoalignment of RNA-seq data using the kallisto algorithm.
Developing these algorithms is not a simple port of the software, but requires
detailed understanding of the hardware, the algorithms and the specific
bioinformatics problem at hand. Given the tremendous parallel processing that
modern GPUs posses, we are expanding our research to include other problems.