Eric W. Klee

A low-rank covariance probe on Evo 2 embeddings predicts ClinVar pathogenicity across SNVs and indels, while annotation probes plus an LLM produce mechanistic hypotheses; scores released for 4.2 million variants.

This study uses long-read sequencing to re-evaluate the pathogenicity of intragenic DMD duplications, finding that many are non-tandem and likely benign.

OmicsFootPrint integrates and interprets multi-omics data using circular images and deep neural networks, achieving high AUC scores in cancer subtype classification and drug response prediction.

The paper identifies AP2S1 variants linked to neurodevelopmental disorders by showing they impair clathrin-mediated endocytosis and disrupt protein interactions.

This study identifies MARK2 variants as a cause of autism spectrum disorder by showing they disrupt the WNT/β-catenin signaling pathway and impair neural development.

UNISOM improves CHIP detection from low-coverage sequencing data by combining variant calling with machine learning classification, identifying low-VAF mutations with high sensitivity.

The paper reports a cohort of 35 individuals with CUL3 variants causing a syndromic neurodevelopmental disorder, showing loss-of-function variants disrupt protein homeostasis by impairing proteasomal degradation of CUL3 substrates.

BOREALIS identifies outlier methylation sites in bisulfite sequencing data from rare disease patients, linking them to phenotypically relevant genes.

This study identifies de novo coding variants in the AGO1 gene as a cause of a novel monogenic neurodevelopmental disorder with intellectual disability and associated behavioral features.

This paper evaluates workflow management systems for genomics research and clinical use, comparing their features and suitability across different environments and use cases.

The paper evaluates the detection of technically challenging variants in NGS workflows using synthetic specimens and clinical data from 450,000 patients, finding significant variability in detection rates.

The paper investigates RNF213 variation's role in neurovascular diseases across Caucasian and Japanese populations, finding linkage disequilibrium between SLC26A11 and RNF213 variants associated with ischemic stroke and aneurysms.

The paper introduces a hybrid variant calling method integrating NGS and TGS reads to improve small variant accuracy, achieving up to 30% reduction in errors in challenging regions.

This study identifies DDX3X mutations as causing neurodevelopmental disorders by disrupting RNA metabolism and neurogenesis during fetal cortical development.

This study compares the accuracy and performance of Sentieon DNASeq and GATK for single-sample variant calling, finding they are nearly identical in accuracy with Sentieon being faster and more scalable.

This study benchmarks GATK3.8 and GATK4 to evaluate performance improvements and cost trade-offs in genomic variant calling.

This study evaluates synthetic specimens with challenging variants across 10 NGS workflows, finding limited variant detection consistency and highlighting bioinformatic limitations.