Kuan‐lin Huang

fastVEP is a fast, comprehensive variant effect predictor implemented in Rust, outperforming the Perl-based Ensembl VEP in speed while maintaining accuracy.

The paper evaluates demographic representation in single-cell omic atlases and finds significant over-representation of European populations and under-representation of Asian and Latino individuals.

LinkD is a drug repurposing platform that integrates molecular, phenotypic, and clinical data to prioritize drugs, showing reduced prostate cancer incidence with beta-blockers.

This paper develops MINDS, a multispecific nanobody-Fc platform that co-depletes membrane receptors and delivers diverse payloads, showing improved antitumor activity in cancer models.

This paper uses an AI agent framework to validate 31 oncology gene target-disease hypotheses, finding non-retracted targets are 17-fold more likely to show context-specific dependency than retracted ones.

The paper investigates tissue-specific chromosomal deletions in BRCA1/2-related cancers using bulk and single-cell data, finding early deletions of 17q and 13q in breast and ovarian cancers.

RastQC is a Rust-based sequencing QC tool that replaces FastQC, offering faster performance, lower memory use, and support for both short and long reads.

The paper identifies genomic deletions as the rate-limiting mutational processes in breast and ovarian cancer initiation, requiring 1-3 deletions for tumor development.

The paper identifies epistatic interactions between cancer driver mutations and allele-specific copy number variations in tumors, showing their impact on patient survival.

The paper analyzes how mutations affect mRNA and protein expression in 953 cancer cases across six types, finding that 47.2% of somatic expression quantitative trait loci (seQTLs) impact protein levels, with some mutations showing disproportionate effects.

The paper identifies Master Protein abundance Regulators (MaPRs) across ten cancer types using a computational pipeline, finding they mediate up to 79% of post-transcriptional regulatory networks and predict tumor proteomic subtypes more accurately.

The paper identifies expression-driven genetic dependency targets for precision oncology using a Bayesian approach, revealing new druggable genes and validating their therapeutic potential.

This study uses Mendelian Randomization to identify causal relationships between circulating biomarkers and diseases, finding novel associations between glucose and cystatin C with bipolar disorder.

The paper introduces HotPho, a tool that identifies 3D co-clustering of phosphosites and cancer mutations on protein structures, revealing 474 hybrid clusters with potential functional significance.

The paper identifies cancer-specific driver gene mutations associated with immunotherapy outcomes in a multi-cancer cohort of 1,525 patients.

The paper identifies ancestry-specific germline variants in cancer predisposition genes and their somatic effects across diverse populations.

This study examines genotype concordance and polygenic risk score differences across three consumer genetic testing services, finding significant discrepancies in Alzheimer's disease risk estimates.

The paper develops ADmiRE to annotate miRNA variants in large population datasets, identifies highly constrained miRNAs, and provides conservation scores for future disease studies.

This study identifies a haplotype lowering PU.1 expression in myeloid cells that delays Alzheimer’s disease onset by affecting gene expression and function.