CPTAC Pan-Cancer PTMs

Repo for manuscript: Geffen, Y., Anand, S., Akiyama, Y., Yaron, T., Song, Y., et al. "Pan-Cancer analysis of post-translational modifications reveals shared patterns of protein regulation." Cell (2023).

[Link]

Contact: Shankara Anand - sanand@broadinstitute.org, Yo Akiyama - yakiyama@broadinstitute.org

---

Repo Structure:

Fig1:

• 01*_process_*: processing files that take raw proteomic data from proteomic data commons and perform filtering, imputation for clustering, and PTM correction for protein abundance.
• 02_process_rna.ipynb: notebook for processing transcriptomic data for all CPTAC cohorts. Transcriptomic data were processed using the GTEx pipeline.
• 03_integrate_omics.ipynb: notebook for integrating RNA, protein, and phosphoprotein for use in ARD-NMF clustering. These are then fed into a script to run SignatureAnalyzer, a bayesian variant of non-negative matrix factorization found here in the scripts folder.
• 04_processing_clustering.ipynb: processing results from SignatureAnalyzer.
• 05_pathways.ipynb: pathway enrichment for each data modality in each signature.
• Fig1ACD: notebook for re-creating Figure 1 A,C,D.

Fig2:

• 01_cluster_samples.ipynb: notebook for clustering CPTAC dataset based on signature weights.
• Fig2A: notebook for creating overview, signature figure.
• Fig2F: violin plots for comparison.

Fig3 (visit directory for details):

• 01_mutational_signatures_extraction.ipynb: Extracting mutational signatures
• 02_signature_thresholding_and_classification.ipynb: Defining mutational signature thresholds for classifying DNA repair deficiencies
• 03_recluster_ddr_deficient.ipynb: Clustering DNA repair deficient cancers with homologous recombination deficiency (HRD) and mismatch repair deficiency (MMRD)
• 04_HRD_Differential_Expression.ipynb: HRD differential expression analysis (global HRD vs HRP and acute vs chronic hypoxia HRD)
• 05_MMRD_Differential_Expression.ipynb: MMRD vs MMRP differential expression analysis
• 06_MRN_complex_analysis_ccle.ipynb: MRN complex proteomic and transcriptomic analysis in CCLE cell lines

Fig4: immune analyses

Fig5: histone analyses

data: see ./data/README.md.

downsampling: folder with downsampling analyses to demonstrate robustness of clustering approach.

funcs: folder with python/R functions used throughout notebooks.

scripts: folder with scripts used throughout analyses.