Michael M. Weinstein

The paper evaluates lysis and extraction methods using metagenomic reference materials to identify optimal techniques and highlights biases in compositional metrics due to poor lysis.

FIGARO chooses read-trimming settings for 16S/18S rRNA gene sequencing data to maximize reads kept after quality filtering in DADA2-based pipelines, reducing trial-and-error.

This paper introduces a machine learning model, Elevation, for predicting CRISPR-Cas9 off-target effects, which outperforms existing methods and provides a new evaluation metric.