Gene-Expression Data Retrieval and Pre-processing

The list of datasets used in this analysis is provided in supplementary tableNeuroblastomaSI.xlsx of the paper.

To retrieve and pre-process gene-expression data:

• For microarray datasets, follow instructions from EMT_score_calculation.
• For RNA-sequencing datasets, refer to EMT_Scoring_RNASeq.

---

Random Gene-Swap Analysis

To explore PC1 variance using random gene combinations:

• Use PC1_Variance_Histogram.r to generate histograms for PC1 variance across random combinations of NOR/MES gene lists and housekeeping genes.

• Add pre-processed gene-expression matrix files as tab-delimited .txt files in the datasets/ folder.

• Use PC1_Swap.r to generate boxplots by swapping NOR/MES genes with housekeeping genes (one at a time).

---

Linear Fit to PC1 Means

To visualize the relationship between gene swaps and PC1 variance:

1. Create a GSEID.csv file in the PC1_Means/ folder with the following structure:

   • Column 1: Number of Swaps
   • Column 2: Mean Variance

2. Run Linear_fit.py to generate the linear fit and obtain:

   • R-squared value
   • Mean squared error
   • Slope and intercept

---

PCA, K-Means Clustering, and GSEA

To perform dimensionality reduction and enrichment analysis:

• Add pre-processed gene-expression matrix files as tab-delimited .txt files to the data/ folder.

• Run PCA.py for Principal Component Analysis and K-Means clustering.

• Use GSEA.py to perform Gene Set Enrichment Analysis on the resulting clusters.

• Provide a gene signature signature.gmt file as input for GSEA. Gene signatures can be obtained from MSigDB.

• Create an Output/ folder to store:

  • PCA plots
  • GSEA results

---

Citation

If you use this repository or its contents in your work, please cite the following publication:

Mutually exclusive teams-like patterns of gene regulation characterize phenotypic heterogeneity along the noradrenergic-mesenchymal axis in neuroblastoma
Cancer Biology & Therapy (2024)
DOI: 10.1080/15384047.2024.2301802