Hello everyone,
I am trying to use Machine Learning approaches to predict cancer outcomes, i.e. to classify samples (patients) according to a certain clinical outcome such as risk, recurrence or survivability.
I am trying to use genomic-based data from The Cancer Genome Atlas (TCGA) and my aim is to come up with datasets in a gene-patient matrix format with the corresponding class labels (for example tumor/normal) for each patient.
I tried to download some breast cancer (BRCA) DNA Methylation data but I am confused as to how to proceed in order to get the matrix in question. So my question is what files should I download and how do process it in order to obtain such a matrix from TCGA data ?
Thank you very much for your help.
0 answers
No answers yet.
Log in to answer this question.
See these past threads for inspiration:
A: How can I download methylation array raw files containing mutation in gene A fro
A: How to exclude some of breast cancer subtypes just by looking at gene expressio
Hi, thanks a lot for your reply. I checked the threads you suggested, but I still can't see how I can get TCGA data in patient-gene matrix format.
If you follow the advice in the first thread, you will be able to produce the data in matrix format by inputting all files into R and then extracting the column for methylation level (Beta value). Doing this, you will also learn many useful skills in data organisation and manipulation.
Good luck.
@Madil83 To get the gene-patient matrix format yo have to do some preprocessing steps
Hello, thanks for your answer! Could you please enlighten me on these preprocessing steps? Thank you.
it depends on your goal. you have to keep only the rows and columns that will help you to get a clean data Preprocessing will include some exclusion or deletion of irrelevant information and values conversion as well.