Peter Lincoln Wang

FLASH predicts phenotype from sequence using deep learning, achieving high accuracy on unseen variants and identifying new virulence predictors across microbes.

sc-SPLASH enables reference-free discovery in scRNA-seq and spatial transcriptomics, uncovering novel transcriptomic features like V(D)J recombination and secreted repeat proteins.

The paper introduces SPLASH and compactors to efficiently discover microbial genetic elements and perform strain typing without reference genomes.

SPLASH is a unifying, statistically driven approach to biological discovery from raw sequencing data, bypassing alignment.

DIVE is a reference-free method for discovering mobile genetic elements and diversity-generating elements, showing improved detection power over existing reference-based approaches.

The paper was withdrawn due to duplicate posting and is no longer available for citation.

SPLASH+ identifies novel transcriptomic diversity in human single cells without reference genomes or cell type metadata.

The paper introduces SpliZ, a statistical method to detect cell-type-specific splicing regulation in single-cell RNA-seq data, finding 9.1% of genes show such regulation in human tissues.