Justin Reese

Systematic benchmarking demonstrates large language models have not reached the diagnostic accuracy of traditional rare-disease decision support tools

This study benchmarks large language models against a traditional diagnostic tool for genetic disease diagnosis using 5,213 case reports, finding LLMs lag behind in accuracy.

KG-Microbe - Building Modular and Scalable Knowledge Graphs for Microbiome and Microbial Sciences

KG-Microbe constructs modular knowledge graphs to integrate microbial data, enabling analysis of complex interactions in health and environment.

GA4GH Phenopacket-Driven Characterization of Genotype-Phenotype Correlations in Mendelian Disorders

The paper introduces GPSEA, a tool using GA4GH Phenopackets to analyze genotype-phenotype correlations in Mendelian disorders, identifying 225 significant associations across 85 cohorts.

Performance of Information Theory Derived Semantic Similarity Algorithms for Differential Diagnosis and Clustering

The paper compares semantic similarity algorithms for differential diagnosis and clustering using HPO, finding Phenomizer performs best for ranking and non-normalized algorithms for clustering.

Replacing non-biomedical concepts improves embedding of biomedical concepts

The paper replaces non-biomedical synonyms with biomedical terms to improve concept embeddings using Word2Vec, reducing mean intracluster distance by 8%.

Intrinsic-Dimension analysis for guiding dimensionality reduction and data fusion in multi-omics data processing

This paper introduces a method to guide dimensionality reduction and data fusion in multi-omics data by analyzing intrinsic dimensionality, showing improved performance in predicting survival outcomes.

AI-readiness criteria for biomedical data

The paper defines AI-readiness criteria for biomedical data, encompassing seven dimensions, and evaluates them across flagship datasets to ensure ethical and explainable AI practices.

A corpus of GA4GH Phenopackets: case-level phenotyping for genomic diagnostics and discovery

The paper presents a large-scale corpus of standardized phenopackets for genomic diagnostics and discovery, curated from literature and useful for software testing and genotype-phenotype analysis.

Leveraging Generative AI to Accelerate Biocuration of Medical Actions for Rare Disease

AutoMAxO uses LLMs to semi-automate MAxO curation for rare diseases, identifying 538 confirmed disease-treatment annotations from 4,918 medical abstracts.

Phenomics Assistant: An Interface for LLM-based Biomedical Knowledge Graph Exploration

Phenomics Assistant is a chat-based interface that uses LLMs to query the Monarch knowledge graph, improving reliability for non-expert users to explore disease-gene-phenotype relationships.

Node-degree aware edge sampling mitigates inflated classification performance in biomedical graph representation learning

The paper introduces a degree-aware edge sampling method to reduce performance inflation in biomedical graph representation learning caused by imbalanced edge degree distributions.

An algorithmic framework for isoform-specific functional analysis

The paper introduces isopret, an algorithm for predicting isoform-specific functions using RNA-seq data, and shows that expression and splicing regulate distinct biological functions.

Supervised learning with word embeddings derived from PubMed captures latent knowledge about protein kinases and cancer

The paper uses word embeddings from PubMed to predict which protein kinases are relevant for treating specific cancers using machine learning.

SvAnna: efficient and accurate pathogenicity prediction for coding and regulatory structural variants in long-read genome sequencing

SvAnna predicts pathogenicity of structural variants in long-read sequencing data by integrating transcript and regulatory annotations with clinical phenotype data, achieving high accuracy in prioritizing pathogenic variants.

KG-COVID-19: a framework to produce customized knowledge graphs for COVID-19 response

KG-COVID-19 creates customizable knowledge graphs for COVID-19 response by integrating siloed biomedical data into structured formats for various research applications.

Interpretable Clinical Genomics with a Likelihood Ratio Paradigm

The paper introduces LIRICAL, a likelihood ratio-based framework for genomic diagnostics that improves accuracy and interpretability in phenotype-driven rare disease diagnosis.