Computational Cancer Biology Resources
Dataset :
Gene Discovery in Prostate Cancer
Radiogenomics of Glioblastoma
Dataset name: The Cancer Genome Atlas (TCGA) GBM dataset with both MRI and gene expression data for analyzing the radiogenomic associations with the survival of each feature.
Dataset link: https://www.cancer.gov/tcga
Description: There are 202 cases and out of them 59 cases have the corresponding MRI data, with T1 contrast and air modalities. The gene expression proles of 59 patients consist of 1740 genes for each patient.
Glioma Survival prediction using mRNA expressions
Dataset name: The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA) cohorts
Dataset link: https://www.cancer.gov/tcga, http://www.cgga.org.cn/download.jsp
Description:
Gene expression GBM and lower grade glioma (LGG), which are biomarkers to classify subjects into risk groups, TCGA dataset comprises of 252 subject cases with overall survival information and gene expression profiles of 16510 genes, obtained using the Illumina HiSeq RNA Sequencing platform.
CGGA dataset consists of 315 cases with overall survival data and gene expression profiles of 24326 genes, acquired using Illumina HiSeq 2000 platform, which is a powerful high-throughput sequencing system.
Techniques and Tools:
Gene Discovery in Prostate Cancer
General Statistics
Binary Classifiers
Support Vector Machines
Radiogenomics of Glioblastoma
Machine Learning algorithms:
Linear Kernel function of Support Vector Machine (linear-SVM),
Radial basis kernel function of Support Vector Machine (r-SVM),
Polynomial kernel function of Support Vector Machine (p-SVM),
Random Forest Classication (RFC),
Decision Tree (DT) and Logistic
Regression (LR)
Glioma Survival prediction using mRNA expressions
Decision Tree
Random forest classification
XGBoost
CatBoost
Support vector machine
Bayesian neural network
Pyro (version 1.3.1) probabilistic programming language
pytorch (version 1.5.0)
Scikitlearn library