Phillip Brooks

The paper identifies five bacterial strains associated with Crohn's disease using a novel method combining machine learning and assembly graphs for metagenomic analysis.

Greedy minimum set cover over FracMinHash k-mer sketches selects a small set of reference genomes explaining a metagenome, supporting species-level taxonomic profiling and choice of genomes for read mapping.

The paper provides strategies for using workflow systems to streamline biological data analysis, emphasizing reproducibility and scalability in high-throughput sequencing projects.

The paper introduces sourmash, a software package that uses MinHash to create compressed sequence signatures for efficient large-scale sequence comparison and database searching.