Gene expression and cancer classification

The paper simulates gene expression compendia to show that technical variability masks biological signals, but correction methods help when few experiments are combined, while more experiments make detection easier without correction.

The paper identifies gene signatures predictive of asthma severity using a denoising autoencoder and random forest classifier applied to sputum gene expression data.

This paper introduces a data-driven method for dimensionality reduction of compositional data using amalgamation, achieving performance comparable to state-of-the-art methods while producing interpretable features.

The paper introduces stingscore, a method that uses stably-expressed genes to score transcriptomic signatures from individual samples, improving normalization and enabling clinical applications.

This paper introduces SampleLASSO, a method for imputing unmeasured gene expression in microarray data, showing it outperforms existing methods across multiple tasks.

The paper introduces AD-AE, an adversarial deconfounding autoencoder that generates gene expression embeddings free of confounders and capable of generalizing across domains.

ComBat-Seq adjusts RNA-Seq count data for batch effects using negative binomial regression, improving statistical power and false positive control in differential expression analyses.

This study refines differential gene expression analysis using machine learning to identify a 10-gene signature for predicting pediatric sepsis mortality with an AUC of 0.89.

The paper applies tensor decomposition-based unsupervised feature extraction to gene expression data from drug-treated mouse tissues, identifying common drug effects across tissue groups.

DataRemix is a normalization method that enhances biological signal detection in gene expression data by tuning hidden factors, outperforming existing methods without external knowledge.

The paper extends the optimal discovery procedure (ODP) to handle complex study designs and RNA sequencing, showing improved performance in detecting differentially expressed genes.

This paper develops a normalization and quality control approach for NanoString RNA data, showing improved handling of technical variation without loss of biological variation.

The paper develops a clinically validated transcriptome test using capillary blood for global systems biology studies, enabling affordable and accessible gene expression analysis.

The paper applies genomic selection methods to multi-omics data to develop prognostic signatures for prostate cancer, showing improved predictability and computational efficiency.

This paper introduces a hierarchical Bayesian mixture model to infer gene expression states from transcriptomic data, showing high accuracy in both simulated and empirical datasets.

dream improves differential expression analysis for repeated measures designs by reducing false positives and increasing power across multiple datasets.

The paper introduces a new information-theoretic distance measure, TSD, to assess transcriptomic similarity between tissue samples using RNA-seq data and curated gene signatures.

This paper applies style transfer methods to RNA-Seq data harmonization and treatment outcome prediction, showing improved style and semantics prediction accuracy compared to existing approaches.

PDXGEM builds a gene expression model to predict cancer patient responses to anticancer therapies using data from patient-derived xenograft models.

The paper introduces TDimpute, a transfer learning-based deep neural network for imputing missing RNA-seq data using DNA methylation, showing improved accuracy across 16 cancer datasets.

The paper introduces CODC, a copula-based model to identify differential coexpression between gene pairs across two conditions, showing improved performance on TCGA RNA-Seq data.

The paper introduces a two-stage framework using mixed mutual information and scalable Bayesian networks to analyze multimodal cancer data, identifying 11 candidate genes with 6 validated in literature.

The paper identifies a diagnostic signature for myocardial infarction using differential expressed genes and validates it on independent datasets.

ClusterMine integrates biological knowledge to cluster samples based on gene set expression profiles, improving interpretability of clusters.

CCLA authenticates cancer cell lines using gene expression profiles with high accuracy and speed, offering a web server for public use.

The paper introduces NBAMSeq, a statistical model for RNA-Seq data that improves detection of nonlinear effects while maintaining linear effect performance.

This paper introduces BESC, a method to correct microarray batch effects without requiring known batch identities or removing true biological differences, using signatures from a reference set.

CAMPP is an open-source R pipeline that simplifies high-throughput data analysis for non-experts, providing tools for normalization, clustering, differential expression, and biomarker prediction.

This paper introduces sparse contrastive principal component analysis to extract stable and relevant biological signals from high-dimensional data and evaluates it on multiple datasets.