Analysis Overview for Head and Neck Squamous Cell Carcinoma
Maintained by TCGA GDAC Team (Broad Institute/Dana-Farber Cancer Institute/Harvard Medical School)
Overview
Introduction

This is the analysis overview for Firehose run "25 July 2012".

Summary

Note: These results are offered to the community as an additional reference point, enabling a wide range of cancer biologists, clinical investigators, and genome and computational scientists to easily incorporate TCGA into the backdrop of ongoing research. While every effort is made to ensure that Firehose input data and algorithms are of the highest possible quality, these analyses have not been reviewed by domain experts.

Results
  • Sequence and Copy Number Analyses

    • Copy number analysis (GISTIC2)
      View Report | There were 288 tumor samples used in this analysis: 23 significant arm-level results, 28 significant focal amplifications, and 50 significant focal deletions were found.

  • Clustering Analyses

    • Clustering of Methylation: consensus NMF
      View Report | The 4669 most variable methylated genes were selected based on variation. The variation cutoff are set for each tumor type empirically by fitting a bimodal distriution. For genes with multiple methylation probes, we chose the most variable one to represent the gene. Consensus NMF clustering of 292 samples and 4669 genes identified 3 subtypes with the stability of the clustering increasing for k = 2 to k = 8 and the average silhouette width calculation for selecting the robust clusters.

    • Clustering of mRNAseq gene expression: consensus NMF
      View Report | The most robust consensus NMF clustering of 229 samples using the 1500 most variable genes was identified for k = 3 clusters. We computed the clustering for k = 2 to k = 8 and used the cophenetic correlation coefficient to determine the best solution.

    • Clustering of mRNAseq gene expression: consensus hierarchical
      View Report | The 1500 most variable genes were selected. Consensus average linkage hierarchical clustering of 229 samples and 1500 genes identified 3 subtypes with the stability of the clustering increasing for k = 2 to k = 8 and the average silhouette width calculation for selecting the robust clusters.

    • Clustering of miRseq expression: consensus NMF
      View Report | We filtered the data to 150 most variable miRs. Consensus NMF clustering of 259 samples and 150 miRs identified 3 subtypes with the stability of the clustering increasing for k = 2 to k = 8 and the average silhouette width calculation for selecting the robust clusters.

    • Clustering of miRseq expression: consensus hierarchical
      View Report | We filtered the data to 150 most variable miRs. Consensus average linkage hierarchical clustering of 259 samples and 150 miRs identified 3 subtypes with the stability of the clustering increasing for k = 2 to k = 8 and the average silhouette width calculation for selecting the robust clusters.

  • Correlation Analyses

    • Correlation between molecular cancer subtypes and selected clinical features
      View Report | Testing the association between subtypes identified by 5 different clustering approaches and 8 clinical features across 290 patients, 5 significant findings detected with P value < 0.05.

    • Correlations between copy number and mRNAseq expression
      View Report | The correlation coefficients in 10, 20, 30, 40, 50, 60, 70, 80, 90 percentiles are 1000.4, 1720, 2331, 2978, 3624, 4282, 4948.8, 5615.2, 6366.6, respectively.

  • Other Analyses

    • Correlation between copy number variations of arm-level result and selected clinical features
      View Report | Testing the association between copy number variation 78 arm-level results and 8 clinical features across 286 patients, one significant finding detected with Q value < 0.25.

    • Correlation between copy number variation genes and selected clinical features
      View Report | Testing the association between copy number variation of 78 peak regions and 8 clinical features across 286 patients, one significant finding detected with Q value < 0.25.

Methods & Data
Input
  • Run Prefix = analyses__2012_07_25

  • Summary Report Date = Fri Aug 17 15:07:18 2012

  • Protection = FALSE