Transcriptomics
Algorithms for RNA-seq quantification, transcriptome analysis, long-read RNA-seq, and downstream inference from abundance estimates.
We build methods and tools for analyzing high-throughput genomics data, with a focus on transcriptomics, single-cell assays, sequence indexing, and scalable inference.
What We Study
Single-cell assays
Scalable workflows for single-cell transcriptomics, feature barcoding, quality control, sparse data, and emerging multimodal sequencing protocols.
Sequence indexing
Compact indexes and mapping methods for large collections of genomes, transcriptomes, and raw sequencing experiments.
Succinct structures
Data structures for de Bruijn graphs, sequence collections, and compressed representations that make large-scale genomics practical.
Software
The lab maintains open-source software used by researchers working with bulk RNA-seq, single-cell RNA-seq, long-read RNA-seq, sequence mapping, indexing, quality control, and graph-based sequence representations.
See the Resources page for curated projects and GitHub-derived metadata.
Lab Approach
Accuracy
We care about methods that support sound scientific inference, not just fast command-line runs.
Performance
We profile, benchmark, and engineer tools for datasets that are large enough to expose algorithmic weaknesses.
Usability
We aim to release software that researchers can install, document, cite, inspect, and build on.
For Prospective Students
COMBINE-lab is part of the Department of Computer Science at the University of Maryland. We generally work with students who are already at UMD and whose interests align with the research areas above.
Most projects in the lab involve algorithm engineering and performance-conscious implementation. New projects are commonly developed in Rust, and new lab members are expected to learn Rust if they do not already know it.
We are usually not the right home for projects centered primarily on deep learning in genomics. We are most interested in students who want to build robust computational methods, data structures, and open software for biological data analysis.
If that sounds like a strong fit, please contact Rob.