/TCGA-survival

Primary LanguageJupyter Notebook

TCGA-survival

Using clinical data and molecular level data to predict survival of cancer patients.

TCGA Pancancer Survival Prediction

2014 data

2014 count GBM KIRC LUSC OV
clinical 210 243 121 379
mRNA 210 243 121 379
miRNA 210 243 121 379
CNV 210 243 121 379
methylation 210 243 x 379
RPPA x 243 121 379
Survival 210 243 121 379
features count GBM KIRC LUSC OV
number of features 43437 38300 21868 43870

2017 data

2017 count OV CESC BLCA ESCA PAAD COAD BRCA CHOL GBM PCPG READ PRAD THCA UCS HNSC KIRP DLBC LIHC LUSC KIRC STAD SARC THYM LGG SKCM TGCT UCEC ACC
Mutation 307 304 408 184 178 191 1093 36 166 179 72 497 501 57 520 290 48 373 501 533 415 262 120 530 103 150 177 79
CNVGene 307 304 408 184 178 191 1093 36 166 179 72 497 501 57 520 290 48 373 501 533 415 262 120 530 103 150 177 79
CNVArm 307 304 408 184 178 191 1093 36 166 179 72 497 501 57 520 290 48 373 501 533 415 262 120 530 103 150 177 79
Protein 307 304 408 184 178 191 1093 36 166 179 72 497 501 57 520 290 48 373 501 533 415 262 120 530 103 150 177 79
mRNA 307 304 408 184 178 191 1093 36 166 179 72 497 501 57 520 290 48 373 501 533 415 262 120 530 103 150 177 79

additional TCGA data (大多為 paired data (Normal, Tumor)

  • Clinical
  • Transcriptome Profiling
  • Biospecimen
  • Simple Nucleotide Variation（4 kinds of variant callers，Somaticsniper, Mutect, Muse, Varscan）
  • Copy Number Variation
  • DNA Methylation

paper

C-index

Kaplan-Meier曲線

Cox model