Supplementary Materials

The PDF file includes:

  • Fig. S1. t-SNE analysis of additional methods to estimate tumor purity in the reference cohort.
  • Fig. S2. Heat map analysis of the top 2000 variable CpG sites from the reference cohort.
  • Fig. S3. Overview of the probability scores of the three machine learning algorithms on the clinical cohort.
  • Fig. S4. Prognostic relevance of the diagnosis predicted by the three different classifiers in the clinical cohort.
  • Fig. S5. Dependency of the cross-validation loss in the reference cohort on the number of CpG sites included in the classification model for the three different classification methods.
  • Table S4. Antibodies used for immunohistochemistry.
  • Table S5. Results from comparative NGS of cases 31 and 43.
  • Legends for tables S1 to S3
  • Legend for data file S1

[Download PDF]

Other Supplementary Material for this manuscript includes the following:

  • Data file S1 contains the following supplementary tables:
  • Table S1 (Microsoft Excel format). Detailed metadata and probability scores for the samples included in the clinical cohort.
  • Table S2 (Microsoft Excel format). Gene enrichment analysis of the 2000 top variable CpG sites of the reference cohort.
  • Table S3 (Microsoft Excel format). Metadata and detailed classification results for LUSC cases from TCGA with previous HNSC.