Module authors: Zhuangzhuang Geng and Jo Lynne Rokita
Note: the previous files/scripts can be find in Archive folder
To run all of the Rscripts in this module from the command line sequentially, use:
bash run-molecular-subtyping-ATRT.sh
run-molecular-subtyping-ATRT.sh
is designed to be run as if it was called from this module directory even when called from outside of this directory.
This folder contains scripts tasked to molecularly subtype ATRT samples in the PBTA dataset.
Methylation classification is used during subtyping.
The DKFZ v12b6 data are available in dkfz_v12_methylation_subclass
and subtypes with dkfz_v12_methylation_subclass_score >= 0.8
are considered high-confidence and used here.
The NIH Bethesda classifier v2 data are available in NIH_v2_methylation_Class
and subtypes with NIH_v2_methylation_Class_mean_score >= 0.9 and NIH_methylation_Superfamily_mean_score >= 0.9
are considered high-confidence and used here.
00-ATRT_subtyping.R
selects samples from histologies-base.tsv
and subtypes all PBTA and/or DGD tumor biospecimens into three subtypes, (ATRT, MYC
, ATRT, SHH
and ATRT, TYR
).
- Filter the samples with
dkfz_v12_methylation_subclass_score >=0.8
and ifdkfz_v12_methylation_subclass
is one of the three ATRT subtypes ->ATRT, MYC
ATRT, SHH
ATRT, TYR
Ifdkfz_v12_methylation_subclass_score
< 0.8 butNIH_v2_methylation_Class_mean_score >= 0.9 and NIH_methylation_Superfamily_mean_score >= 0.9
, andNIH_v2_methylation_Class
is one of the three ATRT subtypes ->ATRT, MYC
ATRT, SHH
ATRT, TYR
- All remaining samples ->
ATRT, To be clasified
Final results is a table with sample_id
, Kids_First_Biospecimen_ID_meth
, Kids_First_Biospecimen_ID_DNA
, Kids_First_Biospecimen_ID_RNA
, and molecular_subtype
, and saved as ATRT-molecular-subtypes.tsv