Depends on the size of the data and number of covariates you want to include, as well as how the data is distributed.

You can use the survival model Jack is talking about whenever you want to estimate the effect of X on time period (e.g. from diagnosis to death, time it takes to get to the ED, time between appointments, etc)

The lasso selection model is used when we have lots of observations and lots of Xs to include as explanatory factors and cannot select which ones are most important

rIght now, Jack is going over how to put together an RNA-Seq dataset from the TCGA to study your cancer type

Ahmad can use this to put together a dataset for renal cancers

Neelaab can use this to put together a dataset of all sarcomas and muscle tumors as well.

We already have LGG dataset for Ikhianosen and GBM data for John

For brain tumors, please get in touch with Jack and me in your respective slack channel to obtain the data.

Everyone can use their slack channel to get further help on how to work with this script if their project involves TCGA.