RNA-seq analysis of Pectobacterium aroidearum SM2 under VOC stress

Code, processed data and results for the transcriptomic analysis in:

Mena Navarro M.P., López Chablé U., Reyes Betanzo C., Arvizu Gómez J.L., Valenzuela Soto J.H., Ramos López M.A., Amaro Reyes A., Hernández Flores J.L., Campos Guillén J. Electrochemical impedance and transcriptomic analysis in Pectobacterium aroidearum SM2 under VOC's-stress during early colonization. Submitted to Microorganisms (2026).

P. aroidearum SM2 was exposed for 1 h to volatile organic compounds (VOCs) from a cell-free filtrate of Kosakonia cowanii Cp1 and compared with an untreated control (two replicates per condition).

A generalized, organism-agnostic version of this workflow is available as rnaseq-bacteria-cookbook (doi: 10.5281/zenodo.20849957).

Repository layout

pectobacterium-voc-transcriptomics/
├── data/
│   ├── metadata_all.csv            # Sample sheet: condition, replicate, batch
│   ├── annotation.tsv              # eggNOG-mapper v2 functional annotation
│   ├── counts/                     # kallisto abundance tables (C1, C2, T1, T2)
│   └── taxonomy/
│       └── reporte_taxonomia.txt   # Kraken2 report (control replicate 1)
├── scripts/
│   ├── preprocessing_linux/        # fastp, Kraken2, MultiQC, kallisto (Bash)
│   ├── 00_PackageInstallation.Rmd
│   ├── 0_TaxonomyReport.Rmd
│   ├── 1_ExploratoryAnalysis.Rmd
│   ├── 2_DataNormalization.Rmd
│   ├── 3_DifferentialExpression.Rmd
│   ├── 4_FunctionalCategorization.Rmd
│   ├── 5_AdvancedAnalysis.Rmd
│   ├── 6_PaperFigures.Rmd
│   └── *.pdf                       # Knitted reports, each ending in sessionInfo()
└── results/
    ├── *.csv, *.pdf                # Count matrices, DESeq2 tables, QC plots
    ├── taxonomy/
    ├── enrichment/                 # GO / KEGG over-representation, pathview maps
    ├── advanced_analysis/          # GSEA, virulence, regulators, PPI network
    ├── TIFFs/                      # 300 dpi versions of the plots above
    └── paper_images/               # Final manuscript figures

How to reproduce

1. Preprocessing (optional). data/counts/ already contains the kallisto output. To regenerate it from raw reads, see scripts/preprocessing_linux/.
2. R packages. Knit scripts/00_PackageInstallation.Rmd once.
3. Analysis. Open paper-pectobaterium-rnaseq.Rproj in RStudio and knit the scripts in numerical order, 1 → 6. Each script reads from ../data and ../results and writes to ../results, so the working directory must be scripts/ (the RStudio default when knitting).

0_TaxonomyReport.Rmd is independent of the others. Scripts 4–6 need internet access (KEGG and STRING downloads).

Script	What it does	Main outputs
1_ExploratoryAnalysis	tximport, low-count filter (row mean > 10), VST, PCA	rawcounts.csv, filtcounts.csv, pca_raw.pdf
2_DataNormalization	ComBat-seq batch adjustment, DESeq2 size factors	adjcounts.csv, normcounts.csv, pca_adj.pdf, boxplots.pdf
3_DifferentialExpression	Wald test, volcano plot, DEG heatmap	deseq_results_Treatment_vs_Control.csv, deg_Treatment_vs_Control.csv
4_FunctionalCategorization	GO and KEGG over-representation, pathview maps	enrichment/, advanced_analysis/full_annotated_results.csv
5_AdvancedAnalysis	GSEA, virulence and regulator profiling, STRING network	advanced_analysis/
6_PaperFigures	Final figures at 1200 dpi	paper_images/

Conventions

• Direction of change. The contrast is Treatment vs Control (contrast = c("Condition", "Treatment", "Control")): a positive log2 fold change means higher expression in VOC-treated cells.
• Differentially expressed genes (DEGs). Adjusted p < 0.01 and |log2FC| > 1. This gives 216 DEGs out of 4,566 tested genes (68 up, 148 down).
• Gene identifiers. Counts are indexed by the RefSeq CDS identifier (lcl|NZ_JBCFOM...cds_WP_xxxxxxxxx.1_n); the WP_ protein accession is the key that links to annotation.tsv.
• Batch. The two replicates of each condition were processed in separate batches (batch column of metadata_all.csv). Counts are adjusted with ComBat-seq and then modelled with design = ~ Condition.

Software

R 4.5.2 with Bioconductor 3.22 (DESeq2 1.50.2, tximport 1.38.2, sva 3.58.0, clusterProfiler 4.18.4, STRINGdb 2.22.0 against STRING v11.5). Preprocessing: fastp 1.1.0, Kraken2 2.17.1, MultiQC 1.33, kallisto 0.51.1. Complete version lists are in the sessionInfo() section of each knitted PDF in scripts/.

Data availability

Raw sequencing reads: NCBI BioProject PRJNA1533106 (BioSample SAMN63298801). Reference genome: P. aroidearum SM2, GenBank JBCFOM000000000.1.

Citation

Please cite the article above. Citation metadata for this repository is in CITATION.cff.

License

Code is released under the MIT License.