SUPPLEMENT: METHODS:
The pathology was reported at Tata Medical Center, Kolkata (LZ and PR).
Review and confirmation of the pathology (CH), and further molecular
testing were performed at the Hospital for Sick Children (SickKids),
Toronto. DNA methylation profiling was done using the Illumina Human
Methylation 850 (850k) array. The idat files were uploaded on the
molecular neuropathology 2.0 website
(https://www.molecularneuropathology.org/mnp) and were analysed on
the brain tumor classifier versions 11b4v.3 and 12.3v.1. Further
analysis, including the unsupervised clustering, was performed in R
v.4.0.3. Unprocessed .idat files for control methylation tumor classes
were downloaded from the NCBI Gene Expression Omnibus (GEO), under
accession number GSE73801, as published before1. The
ChAMP, minfi, limma and Rtsne packages were utilized for downstream
analyses. Raw data were obtained from idat files and normalized using
the single sample normal-exponential out-of-band method (ssNoob). Probes
on the X and Y chromosomes were removed, along with probes with a bead
count of less than 3, those with a detection p value above 0.01, and
probes at known SNP locations or with multiple binding locations as
previously reported by our group.11 T-distributed
stochastic neighbor embedding (t-SNE) analysis using the Rtsne package
applied the following nondefault parameters: theta = 0, pca = F,
max_iter = 7500, and perplexity = 15. Copy number alterations of
genomic segments were inferred from the methylation array data based on
the R-package conumee after additional baseline correction
(https://github.com/dstichel/conumee). Visualization was performed
using the ggplot2 and the plots were edited for aesthetics using Adobe
Illustrator v.25.4.1.