Subtopic Deep Dive

← Gene expression and cancer classification

Differential Expression Analysis
Research Guide

What is Differential Expression Analysis?

Differential Expression Analysis identifies genes with statistically significant changes in expression levels between conditions using statistical models like edgeR and limma.

This subtopic centers on Bioconductor packages such as edgeR (Robinson et al., 2009; 42,430 citations) and methods for multifactor RNA-Seq (McCarthy et al., 2012; 5,580 citations). Researchers apply negative binomial models with empirical Bayes moderation for count data. Over 10 key papers from 2004-2017 address microarray and RNA-Seq applications.

Curated Papers

Key Challenges

Why It Matters

Differential Expression Analysis drives cancer biomarker discovery by pinpointing subtype-specific genes from tumor RNA-Seq data (Robinson et al., 2009). It supports classification of tumor heterogeneity, enabling precision oncology through enriched pathways (Hänzelmann et al., 2013; Chen et al., 2013). In practice, edgeR identifies drivers in multifactor cancer experiments, informing clinical trials (McCarthy et al., 2012).

Key Research Challenges

Multiple Testing Correction

High-dimensional gene expression data requires controlling false discovery rates across thousands of tests. edgeR uses Benjamini-Hochberg adjustment within its negative binomial framework (Robinson et al., 2009). Balancing power and error control remains critical in cancer subtype analyses.

Biological Variation Modeling

RNA-Seq experiments involve complex factors like blocking and replicates, complicating variance estimation. McCarthy et al. (2012) introduced quasi-likelihood methods for multifactor designs. Accurate modeling prevents inflated false positives in heterogeneous tumors.

Transcript vs Gene-Level Inference

Aggregating transcript abundances to gene levels risks losing resolution in differential analyses. Soneson et al. (2015) showed transcript-level estimates improve gene inferences (4,125 citations). This challenges RNA-Seq pipelines for cancer classification.

Essential Papers

<tt>edgeR</tt> : a Bioconductor package for differential expression analysis of digital gene expression data

Mark D. Robinson, Davis J. McCarthy, Gordon K. Smyth · 2009 · Bioinformatics · 42.4K citations

Abstract Summary: It is expected that emerging digital gene expression (DGE) technologies will overtake microarray technologies in the near future for many functional genomics applications. One of ...

WGCNA: an R package for weighted correlation network analysis

Peter Langfelder, Steve Horvath · 2008 · BMC Bioinformatics · 27.5K citations

GSVA: gene set variation analysis for microarray and RNA-Seq data

Sonja Hänzelmann, Robert Castelo, Justin Guinney · 2013 · BMC Bioinformatics · 15.4K citations

Enrichr: a comprehensive gene set enrichment analysis web server 2016 update

Maxim V. Kuleshov, Matthew R. Jones, Andrew D. Rouillard et al. · 2016 · Nucleic Acids Research · 11.0K citations

Enrichment analysis is a popular method for analyzing gene sets generated by genome-wide experiments. Here we present a significant update to one of the tools in this domain called Enrichr. Enrichr...

Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool

Edward Y. Chen, Christopher M. Tan, Yan Kou et al. · 2013 · BMC Bioinformatics · 8.0K citations

Abstract Background System-wide profiling of genes and proteins in mammalian cells produce lists of differentially expressed genes/proteins that need to be further analyzed for their collective fun...

Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation

Davis J. McCarthy, Yunshun Chen, Gordon K. Smyth · 2012 · Nucleic Acids Research · 5.6K citations

A flexible statistical framework is developed for the analysis of read counts from RNA-Seq gene expression studies. It provides the ability to analyse complex experiments involving multiple treatme...

affy—analysis of <i>Affymetrix GeneChip</i> data at the probe level

Laurent Gautier, Leslie Cope, Benjamin M. Bolstad et al. · 2004 · Bioinformatics · 5.3K citations

Abstract Motivation: The processing of the Affymetrix GeneChip data has been a recent focus for data analysts. Alternatives to the original procedure have been proposed and some of these new method...

Reading Guide

Foundational Papers

Start with edgeR (Robinson et al., 2009) for core negative binomial methods, then McCarthy et al. (2012) for multifactor extensions—establishes statistical foundations for 90% of RNA-Seq analyses.

Recent Advances

Study Soneson et al. (2015) for transcript-gene improvements and Kuleshov et al. (2016) for Enrichr updates on DEG enrichment in cancer.

Core Methods

Negative binomial GLM (edgeR), empirical Bayes moderation (limma/edgeR), quasi-likelihood F-tests (McCarthy et al., 2012), probe-level RMA (affy; Gautier et al., 2004).

How PapersFlow Helps You Research Differential Expression Analysis

Discover & Search

Research Agent uses searchPapers with 'edgeR cancer differential expression' to retrieve Robinson et al. (2009; 42,430 citations), then citationGraph reveals 5580-citation extension by McCarthy et al. (2012). findSimilarPapers uncovers GSVA (Hänzelmann et al., 2013) for downstream pathway analysis. exaSearch scans 250M+ OpenAlex papers for limma-edgeR comparisons in tumor subtypes.

Analyze & Verify

Analysis Agent applies readPaperContent to extract edgeR's negative binomial model from Robinson et al. (2009), then runPythonAnalysis simulates count data with NumPy/pandas for FDR validation. verifyResponse (CoVe) cross-checks claims against McCarthy et al. (2012), with GRADE grading statistical methods (A-grade for empirical Bayes).

Synthesize & Write

Synthesis Agent detects gaps like unaddressed transcript-level effects (Soneson et al., 2015), flagging contradictions in microarray vs RNA-Seq. Writing Agent uses latexEditText for methods sections, latexSyncCitations for 10+ papers, latexCompile for biomarker tables, and exportMermaid for edgeR workflow diagrams.

Use Cases

"Reproduce edgeR analysis on TCGA breast cancer RNA-Seq for subtype DEGs"

Research Agent → searchPapers('edgeR TCGA breast cancer') → Analysis Agent → runPythonAnalysis(NumPy/pandas on count matrices, edgeR simulation) → GRADE-verified DEG list with FDR q-values and volcano plot.

"Write LaTeX methods section comparing edgeR vs limma for prostate cancer classification"

Research Agent → citationGraph(Robinson 2009) → Synthesis → gap detection → Writing Agent → latexEditText(methods draft) → latexSyncCitations(5 papers) → latexCompile(PDF with tables).

"Find GitHub repos implementing GSVA for gene set analysis post-DE"

Research Agent → paperExtractUrls(Hänzelmann 2013) → Code Discovery → paperFindGithubRepo → githubRepoInspect(R code) → exportCsv(implementation details for cancer workflow).

Automated Workflows

Deep Research workflow scans 50+ edgeR-related papers via searchPapers → citationGraph → structured report on cancer applications (Robinson et al., 2009). DeepScan's 7-step chain verifies multifactor models (McCarthy et al., 2012) with runPythonAnalysis checkpoints. Theorizer generates hypotheses on WGCNA integration for tumor networks (Langfelder and Horvath, 2008).

Try Doxa for Differential Expression Analysis Research

Frequently Asked Questions

What is Differential Expression Analysis?

It identifies significantly changed genes between conditions using models like edgeR's negative binomial with empirical Bayes (Robinson et al., 2009).

What are key methods?

edgeR for DGE count data (Robinson et al., 2009), quasi-likelihood for multifactor RNA-Seq (McCarthy et al., 2012), and affy for probe-level microarray (Gautier et al., 2004).

What are foundational papers?

edgeR (Robinson et al., 2009; 42,430 citations), WGCNA (Langfelder and Horvath, 2008; 27,507 citations), GSVA (Hänzelmann et al., 2013; 15,439 citations).

What are open problems?

Improving transcript-level inferences for gene summaries (Soneson et al., 2015), modeling biological variation in sparse cancer data, and integrating with enrichment like Enrichr (Chen et al., 2013).