How to apply this tool to my own samples?

[samyangC]

Hi Nok,
its a amazing tool, but some problems occured when i tried to apply it to my own dataset.

In your examples, you used ccle_cellline_pcor_ess_genes.csv as a feature matrix, and it confused me that why you chose to use a similarity matrix instead of a essential genes' matrix?
(if i want to use this tool to predict the drug response of patients in TCGA cohort, how to preprocess the TCGA data?)

Besides, can i use log(AUC) instead of log(IC50)?

[nokcs]

Hi Samyang,

Thank you for your interest. Regarding drug response values, conceptually the model can be trained using any response value, but we have only tested with log2(IC50).

The model does not directly predict drug response based on gene expression values but based on a kernel feature. The kernel feature represents similarity (Pearson correlation across essential genes) between training samples (CCLE or GDSC cell lines) and input samples. This process avoids directly using gene expression values which could be different across batches or platforms, with the assumption that the gene expression correlation between samples is preserved regardless of measurement platforms.

In order to apply the model to your samples, you will need to calculate the kernel feature for the input samples and then apply the model. We are actually working on a new version of CaDRReS, which should be available in two weeks. We will also provide additional Jupyter notebook for step-by-step model training and applying the model to your samples.

[samyangC]

Hi Nok,
Thank you very much for your prompt reply
I quite looked forward to the new version and the tutorials!

[nokcs]

Hi Samyang,

I'm sorry for the very delayed update. You can find the new version of CaDRReS at https://github.com/CSB5/CaDRReS-SC. In notebook directory, there are Jupyter notebooks for training the models and applying the model to a new dataset.

Hope this is still useful for you.