Casey S. Greene

The paper investigates whether common cis-eQTLs explain gene expression deviations from ploidy expectations in Down syndrome using transcriptomic data from the Human Trisome Project and GTEx.

HetNetEX provides exact asymptotic inference for connectivity in heterogeneous biomedical knowledge graphs, outperforming XSwap in speed and accuracy.

The paper introduces a data portal and open-source tools for single-cell transcriptomics of pediatric tumors, providing uniformly processed RNA-seq data and metadata from over 700 samples across 55 cancer types.

This work analyzes CAFA5 submissions, showing improved accuracy in protein function prediction and highlights the role of protein language models and 3D structures in performance gains.

This study shows HGSC transcriptomic subtypes are driven by tumor cellular composition rather than intrinsic biology, with high predictive accuracy using deconvolution methods.

The paper describes the Common Fund Data Ecosystem (CFDE), which integrates NIH Common Fund data resources to enable discovery, analysis, and collaboration across heterogeneous datasets.

This study improves bulk RNA-seq deconvolution by integrating single-cell and single-nucleus datasets through filtering and transformation strategies, achieving near-scRNA-seq accuracy.

The paper deconvolves adipocyte proportions from bulk RNA-seq data of HGSOC tumors and finds higher adipocyte content is associated with worse survival.

This study uses single-cell and spatial transcriptomics to analyze intra- and inter-tumor heterogeneity in ovarian HGSC subtypes, identifying spatial patterns and cell-cell communication differences.

The paper develops mRNA-based classifiers using RNA-seq and machine learning to accurately distinguish bTB-positive from bTB-negative cattle and other pathogens.

This study models the genomic architecture of adiposity and anthropometrics across the lifespan, identifying shared and distinct genetic factors and their associations with health outcomes.

The paper examines how missing cell types in single-cell references affect deconvolution of bulk data and shows that missing cell-type profiles can be detected using non-negative matrix factorization.

BuDDI predicts cell-type-specific perturbations from bulk data by integrating single-cell and bulk profiles using domain adaptation in a variational autoencoder.

Summix2 uses mixture modeling to estimate and adjust for substructure in genetic summary data, improving the robustness and equity of research using diverse datasets.

This paper introduces CCC, a machine learning-based correlation coefficient that detects both linear and nonlinear patterns in gene expression data, outperforming linear-only methods in identifying biologically relevant relationships.

The paper introduces mousiPLIER, a mouse pathway-level information extractor model trained on brain RNA-seq data, which reduces dimensionality and identifies aging-related latent variables.

The paper introduces a network permutation framework to assess how node degree influences edge prediction in biomedical networks and provides an edge prior as a baseline for performance evaluation.

The paper describes the CCPM Biobank, a dual-purpose genomic resource linking health records and genomic data to support research and clinical applications.

This study compares the impact of different optimization methods on model performance and sparsity in transcriptomic prediction using LASSO logistic regression.

This paper introduces PhenoPLIER, a method that maps gene-trait and drug associations into gene modules to reveal disease mechanisms and drug targets.

This study evaluates if DTC canine genetic tests use user-submitted photos to influence breed ancestry predictions, finding one test relies more on photos than DNA.

The paper uses deconvolution on bulk transcriptomic data to show that HGSOC subtypes are driven by tumor microenvironment composition, particularly fibroblast and immune cell content.

OpenPBTA provides a collaborative platform to genomically characterize 1,074 pediatric brain tumors and 22 cell lines, offering analysis modules and somatic datasets for research and clinical applications.

This study evaluates normalization methods for combining microarray and RNA-seq data, finding that quantile and Training Distribution Matching normalization work best for machine learning tasks.

The paper introduces hetnet connectivity search to find meaningful relationships between biomedical entities in heterogeneous networks without requiring supervised training data.

This study evaluates how experimental factors affect the performance of tumor deconvolution algorithms using bulk RNA-seq data.

The paper improves knowledge graph expansion by reusing label functions for multiple edge types in biomedical abstracts.

wenda_gpu provides a fast domain adaptation method for genomic data using GPU acceleration, achieving comparable results to the original wenda while improving cancer mutation prediction on small datasets.

This study compares neural networks and logistic regression for transcriptomic prediction, finding both linear and non-linear signals are present but not always sufficient for neural networks to outperform linear models.

The paper analyzes transcriptional patterns of core and accessory genes in P. aeruginosa strains using a transcriptome compendium, revealing stable core gene expression patterns across strains.