Gene expression and cancer classification

The paper compares binary and rank-based methods for identifying differentially expressed genes using Bayesian statistics on a yeast dataset, highlighting issues with traditional thresholds and advocating for rank-based approaches.

The paper identifies invariant feature genes (IFGs) using a symmetric machine learning approach to distinguish healthy from disease states and finds drug targets in cancer IFGs.

This paper introduces a Bayesian framework for differential gene expression analysis that ranks genes based on statistical evidence rather than binary classification, improving transparency and computational efficiency.

This study identifies biomarker genes and pathways contributing to lung cancer health disparities using a classification framework with SHAP to interpret disparities between racial and sex groups.

This paper introduces JECCA, a method for identifying shared biological signals across gene expression datasets, enabling disease subtyping, biomarker discovery, and cross-species analysis.

The paper predicts plant phenotypic traits using RNA-seq datasets and finds that few hundred samples suffice for tissue classification but thousands are needed for age prediction.

The paper develops a pre-trained transformer model to predict cancer tissue of origin from bulk gene expression data, achieving high accuracy across 32 tissue types.

The paper introduces CPSM, an R package that provides an automated machine learning pipeline for predicting individual cancer patient survival probabilities using multi-omics data.

MAGE identifies aberrantly expressed genes by analyzing entire expression profiles rather than individual genes, outperforming conventional differential expression and density-based clustering methods.

The paper analyzes gene expression dynamics in normal and tumor tissues, identifying N-genes and T-genes that mark transitions to malignancy in prostate, lung, and liver cancers.

This paper adapts a pre-trained transformer model for bulk tumor data to predict patient survival outcomes, achieving strong correlations and outperforming traditional methods.

This study uses classical statistical methods to identify survival-related genes in breast cancer, finding two novel targets and validating results through bootstrapping.

The paper introduces interval assumptions to replace normalization in differential expression and abundance analyses, reducing false positives and negatives through more realistic and flexible assumptions.

The paper identifies sporadic over-activation of co-regulated genes with extreme outlier RNA expression in multiple species and tissues, showing these patterns are not inherited but may reflect gene regulatory network dynamics.

This study evaluates batch correction methods for integrating microarray and RNA-seq data, finding that limma provides the best balance of Type I error control and power for differential expression analysis.

This paper introduces a graph-informed framework that improves gene pathway analysis and disease classification by integrating gene regulatory networks into statistical methods, showing better performance in simulations and a lung cancer dataset.

RCoxNet improves cancer survival prediction by integrating mutation and clinical data with a deep learning Cox model and random walk algorithm, achieving better performance than existing models.

The paper uses machine learning to identify miRNA biomarkers in blood samples for diagnosing micro-invasive breast cancer, finding ten miRNAs with good classification performance.

The paper introduces a de-biased sparse canonical correlation analysis pipeline to identify cancer-related trans-regulated genes from multi-omic data, finding 256 genes associated with estrogen receptor-related CNAs in breast cancer.

The paper develops statistical methods to assess molecular codes' sparsity and efficiency in characterizing cell types across C. elegans and cancer, showing homeodomain proteins and HLA genes do so more effectively than random controls.

The paper shows that a small number of genes can accurately classify tumor types, using a method based on rough set theory.

CGPA provides a platform for multi-context gene-centric prognostic analysis, enabling customized exploration of gene relationships and biomarker discovery in cancer research.

The paper determines optimal sample sizes for bulk RNA-seq experiments in mice by analyzing large-scale comparative studies and shows that increasing sample size reduces false positives and improves discovery rates.

The paper introduces tauX, a gene expression ratio strategy that improves machine learning performance in precision medicine by refining gene expression features.

GeneBag is a cell foundation model trained on single-cell data and fine-tuned on bulk RNA-seq to diagnose and predict cancer prognosis with high accuracy.

The paper uses graph convolutional neural networks with layer-wise relevance propagation to analyze gene expression data and identify genes and pathways associated with BRCAness, outperforming traditional methods.

BulkRNABert uses a pre-trained transformer model on bulk RNA-seq data to improve cancer classification and survival prediction across multiple cancer types.

The paper introduces a gene expression stability metric, the gene homeostasis Z-index, to identify genes under active regulation in different organs and highlights its advantage over traditional mean-based methods.

This study introduces NBSR, a method for differential expression analysis of miRNA-seq data that outperforms existing approaches on simulated and experimental datasets.

The paper combines LASSO regression with a deep neural network to improve lung cancer diagnosis from transcriptomic data, achieving high accuracy and specificity.