Gene expression and cancer classification

The paper evaluates how well cancer cell lines and PDX models recapitulate human tumor and non-diseased tissue gene expression profiles using in silico analysis.

dRFEtools implements dynamic recursive feature elimination to improve prediction accuracy and reduce computational time in omics data analysis.

The paper assesses the basal gene expression of cancer cell lines to determine their relevance to specific human tissues for in vitro toxicology screening.

ABDS provides tools for analyzing biologically diverse samples by addressing missing value imputation, signature gene detection, and expression pattern visualization.

The paper introduces NanoClassifier, a NanoString-based CMS classifier for CRC samples, showing strong accuracy in predicting CMS subtypes from FFPE and FF tumors.

Sagittarius simulates gene expression at unmeasured timepoints using a transformer model, improving understanding of temporal dynamics in biology.

MTM predicts individualized tissue gene expression profiles from peripheral samples using a multi-task learning framework, improving prediction accuracy and preserving biological variation.

The paper analyzes gene expression and clinical traits in mice to find gene-trait associations using ANOVA, finding statistically significant genes linked to clinical outcomes.

The paper transforms omics data from gene-centric to pattern-centric to improve predictive performance and interpretability in TCGA data.

This study develops ASD-cancer, a deep learning framework that identifies survival subtypes in cancers by integrating tumor microbiome and transcriptome data, revealing microbial-gene interactions linked to patient outcomes.

kimma is an R package for linear mixed effects modeling of RNA-seq data, offering improved sensitivity and model complexity compared to existing pipelines.

This study compares transcriptomic profiles of whole blood, plasma-free blood, and serum-free blood samples, finding that plasma-free blood is more similar to whole blood in gene expression and immune characterization.

CONNECTOR analyzes longitudinal tumor growth data to identify clusters and reveal new molecular associations with therapy response.

RoPE improves differential gene expression analysis by using a data-driven adjustment and robust profile likelihood ratio test, showing better performance with larger sample sizes and reliable error control.

LOCC is a tool for visualizing and scoring cutoffs in continuous variables for dichotomous outcomes, showing better performance than ROC curves in prognostic modeling.

E2EGraph predicts prostate cancer pathological stages using gene expression data and graph neural networks, achieving state-of-the-art performance with interpretable marker gene identification.

This study identifies biological and technical sources of expression variation across 49 human tissues using GTEx data, revealing 522 variable transcript clusters with varying confidence in their explanations.

The paper evaluates methods to detect genes with differential expression dispersion in cancer datasets, identifying autophagy-related functions associated with tumor progression.

This study evaluates differential expression analysis in HT-Seq by analyzing p-value distributions across datasets, finding widespread bias and inconsistent results.

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 a method to identify significant gene expression changes in multiple perturbation experiments using knockoffs and applies it to data from the LINCS program.

This paper introduces a new algorithm, NCT, to identify gene signatures for early cancer detection using RNA-seq data from TCGA and GTEx, showing improved classification performance.

The paper develops mechanistic classifiers using biological constraints to improve the robustness and interpretability of gene signatures in cancer prediction.

This paper investigates how batch-effect associated missing values (BEAMs) impact missing value imputation performance and highlights the need for improved strategies.

This study benchmarks RNA-Seq technical parameters in a clinical vaccine trial to establish guidelines for future studies.

The paper uses machine learning on public transcriptome data to predict age and sex from gene expression and identify molecular signatures associated with age and sex differences.

NMFClustering provides GPU-accelerated non-negative matrix factorization for efficient clustering of large RNA-Seq and scRNA-seq datasets.

Percolate combines gene expression with other genomic data to create robust DNA-based predictors of drug response, outperforming individual data types.

CAT PETR is a web application for differential analysis of phosphorylation data from antibody microarrays, offering improved data input and visualization compared to existing tools.

The paper develops a deep learning framework using Concrete Autoencoder to identify key genes for quantifying intratumor heterogeneity, showing that a subset of genes suffices for accurate ITH scoring.