This repo is dedicated to annotation of molecular properties in TCGA and multiple OAC datasets as input to sysSVM
TCGA mutation files: /mnt/lustre/users/k1801782/dataset/mc3_bailey.bed or /home/camp/let/working/Trang/data/mc3.v0.2.8.PUBLIC.code.filtered.maf
TCGA_ASCAT segment files : /mnt/lustre/users/k1801782/dataset/CNV/ASCATforTCGA/segments
TCGA_ASCAT ploidy file : /home/camp/let/working/Trang/data/ascat_tcga_ploidy.txt
TCGA expression files: /home/camp/let/working/Trang/data/TCGA-RNASeq-download16042019/. In which TCGA-RNASeq is the raw RNASeq file downloaded from TCGA portal, TCGA-RNASeq-CODE is the RNASeq for each cancer type (per .rds file)
Script to plot expression of damaged genes for all cancer types: /home/camp/let/working/Trang/scripts/TCGA_AmpHomoLoss.R
Annovar, OncodriveClust and CNV script (for the above result, also in Github): /mnt/lustre/users/k1801782/dataset/Annotation_individual_cancers
Annovar, OncodriveClust and CNV results for each cancer type: /mnt/lustre/users/k1801782/dataset/Annotation_individual_cancers
Inside each folder (for each cancer type) you will find the following main files:
ANNOVAR/muts_ann_damaging.rds : annotated mutations, and damaging classification according to function, splicing and conservation
OncodriveClust/muts_ann_onco_damaging.rds: annotated mutations (including damaging classification) and OncodriveClust results for individual cancer type
OncodriveClust/muts_ann_onco_damaging_pancan.rds: annotated mutations (including damaging classification) and OncodriveClust results for all cancer types
CNV/CNVGainLoss.rds: results of the intersection between ASCAT segments and hg19 gene coordinates, with CNV_type_corrected
total_table.rds: table counting number of truncating, non truncating damaging mutation, gain of function, CNVGain and Loss for each gene in each patient, based on Annovar and OncodriveClust (individual cancer) annotation
total_table_pancan.rds: table counting number of truncating, non truncating damaging mutation, gain of function, CNVGain and Loss for each gene in each patient, based on Annovar and OncodriveClust (pan cancer) annotation
Combination of files and filtering (for 34 cancer types):
Allcancertypes_individual_annotation_damaging.rds: total_table of 34 cancer types with only damaging mutations (truncating, non truncating damaging mutation, gain of function, CNVGain and Loss), based on ANNOVAR, OncodriveClust (individual cancer) and CNV annotation
Allcancertypes_pancan_annotation_damaging.rds: total_table of 34 cancer types with only damaging mutations (truncating, non truncating damaging mutation, gain of function, CNVGain and Loss), based on ANNOVAR, OncodriveClust (pan cancer) and CNV annotation
Allcancertypes_pancan_annotation_damaging_falseHomotoHeteroDel.rds: total_table (annovar, oncodriveclust and CNV) with damaging mutations after removing false positive Homozygous deletion (genes with CN = 0 but expression (RNA_seq) > 1/cancertype_purity), and move those with a damaging mutation to Heterozygous deletion (change CN ==1) (This is the final file to use, but if you want to filter for different expression threshold, then use the above 2 files)
Other files that might be important/interesting:
/mnt/lustre/users/k1801782/dataset/TCGA_2015_totaltable: Total table for 31 TCGA cancer types annotated by Thanos in 2015 (based on individual files in /mnt/lustre/users/k1469280/mourikisa/data/TCGA/01_03_2015)
/mnt/lustre/users/k1801782/dataset/TCGA_2015_totaltable/total_table_pancan_damaging.rds: Total table with only damaging mutations for 31 TCGA cancer types annotated by Thanos in 2015
/mnt/lustre/users/k1801782/dataset/TCGA_2015_totaltable/total_table_pancan_damaging_withpurity.rds: Total table with only damaging mutations for 31 TCGA cancer types annotated by Thanos in 2015, removing samples without purity values
Softwares: /mnt/lustre/users/k1801782/softwares or /home/camp/let/working/Trang/softwares