Replicating Adebari et al Paper Results

TCGA

1. Download the PPI file from GRAND.
2. Register for a WebMeV account and download the GBM and LGG data from WebMeV.
3. Download the curated gene set GMT file from Human MSigDB.
4. Download the protein coding genes from GENCODE (basic gene annotation).
5. Run separate_samples.py to separate data into male and female and find overlapping genes, changing the following:
   • gbm_file_path: The path to the GBM expression data.
   • lgg_file_path: The path to the LGG expression data.
   • split_file_dir: The path to the directory where you wish to store the split files.
6. Run concatenate.py to concatenate together the GBM and LGG files.
7. Run NORMALIZED_DATA_CODE.r to normalize the data, changing the following:
   • concatenatedFile: The path to the concatenated file.
   • normalizedFile: The path where you wish to save the normalized file.
8. Run Code_to_filter_by_PCG.r to filter by protein coding gene, changing the following:
   • mappingFile: The path to the protein coding genes.
   • matrixFile: The path to the concatenated data.
   • filteredMatrixFile: The path where you wish to save the filtered data.
9. Run updatecode.py to split the normalized and filtered data into LGG_F, LGG_M, GBM_F, and GBM_M.
   • split_file_dir: The path to the directory where you wish to store the split files.
   • matrix_dir: The path to the file containing the matrix of gene expression levels.
10. Run runpanda.py to run PANDA, changing the following:
    • localDir: The local directory where the motif, PPI, and expression data are stored.
11. Run PercentileMonster.R to run MONSTER, changing the following:
    • localDir: The local directory where the files are stored.
12. Run diffanalysis.R to do pathway analysis changing the following:
    • localDir: The local directory where the files are stored.
    • pathwayDBFile: The pathway file path.
13. Run PathwayDifferentialAnalysis.R to plot the pathways, changing the following:
    • sourceDir: The local directory where the files are stored.
14. Run RunBLOBFISHAllTFS.R to run BLOBFISH and plot the results, changing the following:
    • sourceDir: The local directory where the files are stored.

REMBRANDT (GSE108474)

1. Download the clinical and gene expression data from GEO and unzip them.
2. Run preprocessREMBRANDT.r to normalize the data, separate samples into male and female GBM and LGG, and limit to protein coding genes, changing the following: - sourceDir: The path to the REMBRANDT gene expression and clinical data. - pccFile: The path to the file containing gene annotations, to obtain protein coding genes
3. Run runPANDA_REMBRANDT.R to run PANDA on each subgroup, modifying the following:
   • fileDir: The path to the REMBRANDT expression files from step 2
4. Run BLOBFISH on REMBRANDT data using blobfishforrevdata.r, changing the following:
   • sourceDir: The path to the REMBRANDT PANDA files.
   • monsterDir: The path to the MONSTER files from TCGA.
   • pathwayFile: The path to the gene-pathway mapping (GMT format).
   • nullDir: The path to the null PANDA files.
5. Run plotBLOBFISH_REMBRANDT_noAR.R to plot the BLOBFISH results, changing the following:
   • pathwayDir: The path to the TCGA pathway results.
   • rembrandtResultDir: The path to the REMBRANDT BLOBFISH results.
   • rembrandtDir: The path to the REMBRANDT gene expression and clinical data.
   • gmtFilePath: The path to the file containing mappings from genes to pathways (GMT format).
6. Run compareBLOBFISH_HIF1A.R to compare the targeting of specific genes within pathways downstream of HIF1A between male and female GBM in TCGA and REMBRANDT, changing the following:
   • pathwayDir: The path to the TCGA pathway results.
   • tcgaDir: The path to the TCGA gene expression and clinical data.
   • rembrandtDir: The path to the REMBRANDT gene expression and clinical data.
   • gmtFilePath: The path to the file containing mappings from genes to pathways (GMT format).