The second homework is about a multi-omic analysis of brain tumors called gliomas using mutations and copy-number-variations along with other clinical variables.
With this hands-on exercise, I first ask you to do some exploratory data analysis.
Then I ask you to carry out a manual hyperparameter search for classifying cancer types in the dataset.
Finally, I ask you to carry out a single task and multi-task automated hyperparameter tuning to predict survival risk scores of patients.
Please download the following notebook:
wget https://raw.githubusercontent.com/BIMSBbioinfo/compgen_course_2025_module3/refs/heads/main/homeworks/hw2/day2_hw_lgg_gbm.ipynb
I provide some chunks of code and text that are required for subsequent questions.
I left some cells as blanks which I ask you to fill in as part of exercises.
If you have questions or run into issues, please use the appropriate category of the discussion board.