Figure 1

MS-MIMIC, a novel assay to assess minimal DNA methylation signatures, and its application in a proof of concept study for the identification of molecular disease subgroups in medulloblastoma. (a) Summary of assay design and development including derivation and validation of minimal DNA methylation signatures. (b) Non-negative matrix factorisation (NMF) consensus clustering identified four molecular subgroups (WNT, SHH, Grp3 and Grp4), defined by four metagenes. (c) Assessing MS-MIMIC performance against 450k DNA methylation microarray using 101/106 independent medulloblastoma samples. 5/106 samples failed QC (>6/17 CpG locus fails). (d) Subgroup classification from MS-MIMIC showed total concordance with the reference subgroup (from 450k and GoldenGate DNA methylation arrays and CTNNB1 mutation status) after applying a classification confidence probability threshold (NC –non classifiable). (e) Estimates of methylation (β-values) using MS-MIMIC assay correlated closely with gold-standard 450k DNA methylation microarray (n = 91). (f–h) Subgroup assignment and probability estimates (dots) along with 95% confidence intervals (boxplots) for 101 medulloblastoma samples for which DNAs were derived from three types of materials: Fresh frozen biopsies (n = 40), formalin-fixed paraffin-embedded biopsies (FFPE, tumour section; n = 35) and FFPE-derived cytospin nuclear preparations (n = 26). Samples which did not exceed a classification confidence probability threshold of 0.69 (red-line; empirically derived, Supplementary Fig. 5) were deemed nonclassifiable.