Hi,
I would like to ask if any one of you has explored the TCGA database and found any data sets that contains the following information:
1. treatment that patient received
2. gene expression data of patient's tumor
3. if the patient is still alive/ died after x number of days
Why I am asking is because I would like to do a survival analysis on patients treated with a certain drug expressing high/ low levels of a certain gene.
Thank you very much in advance.
For point 2 and 3 the TCGA information is available. For point 1 you may need to look at the patients data for the treatment you are interested in.
Dear Albino,
Thank you very much.
Does that mean if I want to only pick patients that were treated with a certain drug before, I would need to go through individual patient's data to check?
Stephen,
The survival analysis that you mentioned is bit tricky with the data available.
1) Majority of samples considered for RNA-seq in TCGA are obtained from patients who have either not received the treatment or no information is available regarding treatment. [For example, in case of breast cancer dataset, only 127 patients out of 1098 have received prior adjuvant treatment]
2) But if you check followup data, you see treatment information for more patients.
In case of most samples you cannot assess the influence of drug treatment on gene expression & patient survival.
Hello Darshan,
The TCGA set of data fits what I need, actually. I want to see if the overexpression of a certain gene affects the survival outcome of a certain drug.
The pre-treatment RNA seq information is good for me, as well as the survival of the same group of patients treated with the drug of interest afterwards. The problem now is for me to extract a group of patients that were treated with the drug I am interested in.
Guess I will have to find a way round.
Thanks for your input!
I know of this article (https://bioconnector.org/r-survival.html) but I have not used it and I do not know if it may be useful for you. As mentioned above, it seems that you will need to sift through the data carefully to make sure you have the info you need. For each cancer dataset there are several biospecimen and clinical files, and you might need to go through them, possibly through each patient, to look for what you need. In general, if possible I prefer to attempt verifying my myself the info.
Thank you very much, Albino. Will give that package a check.
Thank you all so much for the help! :)
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs