Standardizing RNA-seq Analysis of Fungal Pathogens Using BRC-Analytics

A Candidozyma auris case study demonstrating reproducible RNA-seq analysis workflows.

Overview

This repository contains the manuscript, data, and supporting materials for a white paper demonstrating the utility of BRC-Analytics for standardized fungal pathogen RNA-seq analysis.

Key findings:

• Re-analyzed RNA-seq data from two high-impact publications (Santana et al. 2023 Science, Wang et al. 2024 Nature Communications)
• Achieved strong correlation with published results using official NCBI gene ID mapping (Santana R² = 0.89-0.94, Wang R² = 0.98-0.9998)
• Demonstrated that standardized workflows with versioned references enable reproducible analyses
• Documented an AI mistake: LFC correlation mapping appeared successful (R² = 0.9996) but was only 1% accurate—corrected using NCBI old_locus_tag mapping validated by protein sequence identity

Repository Structure

whitePaper2/
├── MANUSCRIPT.md          # Main manuscript (Markdown with Pandoc citations)
├── MANUSCRIPT.pdf         # Compiled PDF with figures
├── compile.sh             # Pandoc compilation script
├── references.bib         # BibTeX bibliography
├── vancouver-brackets.csl # NIH numbered citation style
├── Figure1_combined_overview.png    # Literature survey overview
├── Figure2_combined_analysis.png    # Standardization challenges
├── Figure3_santana_validation.png   # Santana et al. validation
├── Figure4_wang_validation.png      # Wang et al. validation
├── Cauris_SRA.csv         # NCBI SRA metadata for C. auris
├── STEPS.md               # Complete research/writing documentation
├── CITATION_GUIDE.md      # How to use the citation system
├── REFERENCES.md          # Organized reference list
└── logical_outline.md     # Original manuscript outline

Quick Start

Compile the Manuscript

# Generate PDF (default)
./compile.sh

# Generate Word document
./compile.sh docx

# Generate HTML
./compile.sh html

# Generate all formats
./compile.sh all

Requirements

• Pandoc (>= 2.9): sudo apt install pandoc pandoc-citeproc
• LaTeX (for PDF): sudo apt install texlive-latex-base texlive-latex-extra texlive-fonts-recommended

Manuscript Summary

Abstract

Candidozyma auris has emerged as a critical global health threat due to multidrug resistance and healthcare-associated transmission. We demonstrate the utility of BRC-Analytics combined with agentic AI (Claude Code Agent) for reproducible RNA-seq analysis. By re-analyzing data from two recent publications using a defined reference genome (GCA_002759435.3), IWC workflows, and official NCBI gene ID mapping, we achieved strong correlation with published results. We also document a cautionary tale: an AI-proposed LFC correlation mapping method appeared successful (R² = 0.9996) but was only 1% accurate, highlighting the critical importance of validating AI outputs against authoritative sources.

Key Statistics

• Literature survey: 20 published C. auris RNA-seq studies with linked BioProjects (2018-2025)
• SRA analysis: 27,201 runs across 237 BioProjects; 27% of projects are RNA-seq
• Validation (using NCBI official old_locus_tag mapping):
  • Santana et al.: R² = 0.89-0.94 (165-203 DEGs mapped), 97-99% direction agreement
  • Wang et al.: R² = 0.98-0.9998 (76-259 DEGs mapped), 100% direction agreement

Data Sources

Validation Studies

Study	Journal	BioProject	Focus
Santana et al. 2023	Science	PRJNA904261	SCF1 adhesin, biofilm formation
Wang et al. 2024	Nature Comm	PRJNA1086003	Glycan-lectin interactions

Galaxy Histories (Reproducibility)

• Santana et al.: https://usegalaxy.org/u/cartman/h/prjna904261-final
• Wang et al.: https://usegalaxy.org/histories/view?id=bbd44e69cb8906b58b85fc3ebc05b72b

Analysis Scripts

Full analysis code and intermediate files available at:

• https://github.com/nekrut/claude-projects/tree/main/rnaseq/Cauris_rna_seq_survey
• https://github.com/nekrut/claude-projects/tree/main/rnaseq/santana24_PRJNA904261
• https://github.com/nekrut/claude-projects/tree/main/rnaseq/wang24_PRJNA1086003

Citation Management

This manuscript uses Pandoc with BibTeX for citations (similar to LaTeX).

Adding Citations

In MANUSCRIPT.md:

First described in 2009 [@satoh2009], *C. auris* has spread worldwide.
Multiple studies [@santana2023; @wang2024] demonstrate...

In references.bib:

@article{satoh2009,
  author = {Satoh, Kazuo and others},
  title = {Candida auris sp. nov.},
  journal = {Microbiology and Immunology},
  year = {2009},
  doi = {10.1111/j.1348-0421.2008.00083.x}
}

See CITATION_GUIDE.md for full documentation.

Figures

Figure	Description	File
1	Literature survey overview (32 studies, tool consensus)	Figure1_combined_overview.png
2	Standardization challenges (genome versions, gene IDs)	Figure2_combined_analysis.png
3	Santana et al. validation (R^2 scatter plots, SCF1)	Figure3_santana_validation.png
4	Wang et al. validation (in vitro/in vivo, key genes)	Figure4_wang_validation.png

Methods Summary

1. Literature Survey: NCBI GEO + PubMed/Europe PMC searches identified 20 RNA-seq studies with linked BioProjects
2. SRA Analysis: Analyzed 27,201 runs across 237 BioProjects for C. auris (NCBI:txid498019)
3. Reference Genome: C. auris B8441 GCA_002759435.3 (5,593 genes)
4. Pipeline: FastQC -> fastp -> STAR -> featureCounts -> DESeq2 (IWC workflows)
5. Gene ID Mapping: Official NCBI old_locus_tag attribute to reconcile annotation versions (v2 → v3)
6. Validation: Protein sequence identity (100% match) confirms correct gene correspondence

Status

• Literature survey and SRA analysis
• Re-analysis of Santana et al. (2023)
• Re-analysis of Wang et al. (2024)
• Manuscript draft with figures
• Citation system setup
• User to write "Obtaining Data from BRC-Analytics" section

License

This work is provided for academic and research purposes.

Acknowledgments

• BRC-Analytics - Pathogen bioinformatics platform
• Galaxy Project - Workflow execution
• IWC - Intergalactic Workflow Commission

---

Manuscript generated with Claude Code