COMBINE-lab

Home

The COMBINE lab was created in the fall of 2014, when Rob Patro joined the Department of Computer Science at Stony Brook University. In 2019, the lab moved to the Department of Computer Science at the University of Maryland. Our research interests span many areas of computational / algorithmic genomics, but our core focus is on the development of algorithms, data structures, and statistical inference methods for analyzing high-throughput sequencing data.

Research Interests

The research interests of our lab broadly span many areas of computational biology and bioinformatics. Some of our main areas of interest include:

  • Algorithms for high-throughput transcriptomics (RNA-seq)
    • assembly, quantification, de novo analysis, differential expression
    • gene expression analysis of single-cell sequencing data (specifically, single-cell RNA-seq)
    • analysis of transcriptomes using long-read RNA-sequencing technologies (PacBio and ONT)
    • enhanced modalities such as spatial single-cell transcriptomics and related assays such as single-cell ATAC-seq
  • Data structures for indexing genomes and raw genomic data
    • data structures for building, indexing and querying one or more reference genomes
    • data structures for indexing collections of tens of thousands of raw sequencing experiments
    • fundamental improvements to succinct data structures and their applications to genomics
  • Applications of approximate and data-driven representations and data structures in bioinformatics

For interested students

Are you interested in working in the COMBINE lab? That’s great! However, there are a few considerations that you should make before reaching out.

While individual exceptions do exist, we generally only directly work with students who are already students at the University of Maryland.

Furthermore, please note the research interests listed above. While they are not comprehensive, and we are happy to consider projects that align with our general goals, we typically do not have the resources or desire to pursue otherwise interesting projects that are far afield of these interests. In particular, we are unlikely to accept students or support projects that focus on topics like deep learning in genomics or transcriptomics — there are plenty of other groups that are working on those topics.

Finally, we care deeply about algorithm engineering and the overall performance and reliability of the algorithms and data structures we design and implement. Thus, it is expected that most members of the lab will either know, or be willing to learn a language relevant to such performance concerns. Most of our new projects are developed using Rust, and new members of the lab are expected to learn Rust if they don’t already know it.

If, considering the above, you’re still interested (or, hopefully, even more interested) in joining the lab, please reach out to Rob!