MOESM4 of Deconvolution of transcriptomes and miRNomes by independent component analysis provides insights into biological processes and clinical outcomes of melanoma patients

Additional file 4: Table S1. Survival data of TCGA SKCM covering patient ID, age at diagnosis, year of diagnosis, disease status at last contact, days to death, vital status, days to last contact after initial diagnosisdiagnosis, year of last contact, tumour free survival, survival in years and age at death. Data items have been either directly extracted from individual XML-files holding clinical information publicly available at GDC or calculated based on the extracted information. Table S2. Clinical data of TCGA SKCM covering patient ID, gender, sample type (primary tumour and metastatic) and publication based data for RNA-seq cluster (immune / keratin / MITF-low). Table S3. Parameters of clinical samples and controls in the investigation dataset. Table S4. Ct values obtained for the new samples using qPCR arrays. During analysis, all absent (NA) values and Ct > 36 were replaced by 36, value selected as detection limit. Table S5. Summary of the ICA results for mRNA data. Stability of each component after 1000 runs, results of the survival analysis, number of genes significantly involved (adj.p-value< 0.01) and number of enriched GO biological processes (adj.p-value< 0.01) are reported. Lower and higher estimates of log hazard ratio (LHR) correspond to 95% C.I. Table S6. Summary of the ICA results for miRNA data. Stability of each component after 1000 runs, results of the survival analysis and number of significantly involved miRNAs (adj.p-value< 0.01) are reported. (XLSX 122 kb)