Eric Banks

The paper presents a large reference panel of 515,579 genomes to improve rare-variant imputation across underrepresented populations, enhancing accuracy for variants with frequencies as low as 0.002%.

The paper analyzes over 800,000 human genomes to assess penetrance of clinically relevant variants, finding that 95% of pLoF variants in severe diseases have explanations for lack of disease manifestation.

blend-seq improves variant discovery by combining short and low-coverage long reads, enhancing SNP detection and structural variant recall while maintaining precision.

This study identifies shared and distinct ultra-rare genetic risk factors for epilepsy subtypes using whole-exome sequencing of 20,979 patients and 33,444 controls.

Rare-variant depletion in 1kb windows of 76,156 gnomAD genomes, versus feature-adjusted mutation expectations, marks regulatory elements and outperforms seven alternative scores at flagging functional non-coding variants.

The paper develops a method to infer compound heterozygosity from exome sequencing data, achieving high accuracy in phasing rare variants for recessive disease diagnosis.

CellBender removes systematic background noise from droplet-based single-cell data using a deep generative model, improving alignment with established gene expression patterns.

This paper introduces MAS-ISO-seq, a method to boost throughput of long-read RNA sequencing for isoform discovery, showing improved detection of differentially spliced genes and cell clustering in T cells.

GATK-gCNV detects rare copy number variants from exome sequencing data, achieving high recall in benchmarking and generating a reference catalog of rare coding CNVs in the UK Biobank.

Open flow cell on a circular wafer with mostly natural nucleotides and no reversible terminators: about 300 bp reads at $1/Gb, 99.6% SNP accuracy on reference human genomes.

The paper presents a pipeline to call mitochondrial DNA variants in gnomAD, reporting population frequencies for 10,850 unique mtDNA variants across 56,434 individuals.

The paper describes AnVIL, a cloud-based platform that enables genomic data analysis without data transfer, improving security, scalability, and collaboration.

The paper evaluates potential drug targets using loss-of-function genetic variants from gnomAD, highlighting the importance of deep curation and targeted recall studies for drug discovery.

From 125,748 exomes and 15,708 genomes, 443,769 high-confidence predicted loss-of-function variants yield LOEUF, a continuous gene intolerance score tracking mouse lethality, cell essentiality, and disease associations.

The paper develops an expression-based annotation metric to improve rare variant interpretation by filtering false-positive loss-of-function variants and identifying pathogenic variants in highly expressed exons.

Short-read genome sequencing of 14,891 diverse genomes yields 433,371 structural variants (SVs) for gnomAD; loss-constrained genes also resist duplication, and clinical SV carrier rates are benchmarked.

The paper analyzes loss-of-function variants in LRRK2 across large genomic datasets to assess the safety of partial LRRK2 inhibition as a Parkinson's disease therapy.

The paper systematically studies uORF variants in 15,708 genomes, showing they are under strong negative selection and contribute to disease.

The paper analyzes multi-nucleotide variants (MNVs) in 125,748 exomes and 15,708 genomes, identifying their distribution, mutational origins, and functional implications.

The paper introduces a benchmark dataset of validated somatic mutations derived from cell lineage, providing a large collection of true somatic variation for use in mutation discovery pipelines.

The paper defines functional equivalence standards for genome sequencing pipelines to enable harmonized variant calling across diverse datasets with minimal batch effects.

Per-sample assembly with gVCF output, followed by joint genotyping, lets GATK HaplotypeCaller scale linearly with cohort size while calling indels more accurately than SAMtools, UnifiedGenotyper or Platypus.

Four levers for cheaper cloud genomics on FireCloud: size virtual disks dynamically, tune VM provisioning from monitoring data, buy preemptible machines, and parallelise tasks at the right width.

This paper analyzes protein-coding genetic variation in 60,706 humans to catalog diversity, assess pathogenicity, and identify genes under strong selection.