Shaista Madad

MetagenomicsBench evaluates AI agents' ability to analyze metagenomic data through 100 verifiable tasks, finding that performance varies across domains and that technical plausibility does not guarantee correct biological interpretation.

Human Developmental Cell Atlas integrates ~4.6 million cells/nuclei from 4-22 post-conception weeks into 454 refined cell types and maps 114 spatial niches across 6PCW whole-embryo sections.

Dictionary learning enables assumption-free cross-modality single-cell mapping via multi-omic bridge datasets and scales single-modality integration to millions of cells in under an hour via atomic sketching.

Signac extends Seurat with a ChromatinAssay for single-cell chromatin and multimodal analysis, adds modified TF-IDF weighting, peak-gene linkage and mitochondrial-variant clonotyping, and scales to 733,526 cells.