mosGraphGen

If you want to cite this tool, please add the citation of following paper link:

https://www.biorxiv.org/content/10.1101/2024.05.15.594360v1.full.pdf

Or contact me via email: hemingzhang@wustl.edu

1. UCSC Dataset Processing

• Parse the data from the UCSC Xena website in PANCAN cohort: https://xenabrowser.net/datapages/?cohort=TCGA%20Pan-Cancer%20(PANCAN)&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443

• Copy number - somatic mutation (SNP and INDEL) - Gene level non-silent mutation

  • Dataset: https://xenabrowser.net/datapages/?dataset=mc3.v0.2.8.PUBLIC.nonsilentGene.xena&host=https%3A%2F%2Fpancanatlas.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443

• DNA methylation (Methylation450K)

  • Dataset: https://xenabrowser.net/datapages/?dataset=jhu-usc.edu_PANCAN_HumanMethylation450.betaValue_whitelisted.tsv.synapse_download_5096262.xena&host=https%3A%2F%2Fpancanatlas.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443
  • ID Map: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GPL16304

• Gene expression RNAseq - TOIL RSEM fpkm

  • Dataset: https://xenabrowser.net/datapages/?dataset=tcga_RSEM_gene_fpkm&host=https%3A%2F%2Ftoil.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443

• Protein expression - RPPA

  • Dataset: https://xenabrowser.net/datapages/?dataset=TCGA-RPPA-pancan-clean.xena&host=https%3A%2F%2Fpancanatlas.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443

• Phenotype - Curated clinical data

  • Dataset: https://xenabrowser.net/datapages/?dataset=Survival_SupplementalTable_S1_20171025_xena_sp&host=https%3A%2F%2Fpancanatlas.xenahubs.net&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443

Check the jupyter nodebook 'UCSC_union_raw_data_process.ipynb' for details.

2. ROSMAP Dataset Processing

• parse the multiomics data from the synapse website of the dataset ROSMAP https://www.synapse.org/#!Synapse:syn23446022

• Genome Variants https://www.synapse.org/#!Synapse:syn26263118

• Methylation https://www.synapse.org/#!Synapse:syn3168763

• RNA sequence https://www.synapse.org/#!Synapse:syn3505720

• Proteomics https://www.synapse.org/#!Synapse:syn21266454

• parse the clinical data https://www.synapse.org/#!Synapse:syn3191087

Check the jupyter nodebook 'ROSMAP_union_raw_data_process.ipynb' for details.

3. Run the Graph Neural Network Model

3.1 Load the data into NumPy format

change the dataset arguments if different dataset will be selected to run the experiments.

python load_data.py --dataset 'UCSC'

3.2 Run experiments on UCSC dataset

Select the fold_number to run individual split of the dataset. Numbers ranging from 1 to 5 here are allowed since the 5-fold cross validation is used.

python geo_UCSC_tmain_gcn.py --fold_number 1

python geo_UCSC_tmain_gat.py --fold_number 1

python geo_UCSC_tmain_gin.py --fold_number 1

python python geo_UCSC_tmain_gfomrer.py --fold_number 1

3.3 Run experiments on ROSMAP dataset

python geo_ROSMAP_tmain_gcn.py --fold_number 1

python geo_ROSMAP_tmain_gat.py --fold_number 1

python geo_ROSMAP_tmain_gin.py --fold_number 1

python geo_ROSMAP_tmain_gformer.py --fold_number 1

4. Downstream analysis for with patient specific biomarkers discovery

The R programing language will be used here to visualize the result in the file 'UCSC_gformer_vis.R', with the attention mechanism in model GraphFormer.

Following is an patient-specific biomaker discovery analysis exmaple. For patient TCGA-D7-6525-01