If possible, it would be nice to have some elaboration on some of the data files and how they align in time. 1) Are all the clinical data obtained at the same time as the genomic and dose-response data, or are they possibly from diagnosis or inclusion? 2) Are survival times used as response for sc2 given as time from diagnosis, inclusion or latest biosample (corresponding to genomic and dose response data)? 3) Are all patients treated with the same regimen? Thanks.

Created by Rasmus Brøndum rfbrondum
Hi Olivier, The drug will certainly affect the RNA profile! The distinction you're making is a good one: our dataset includes RNA profiles of samples *before* drug administration, not after. There exist other datasets which measure RNA profiles after drug administration, and they might have a different RNA profile for every (drug, concentration) pair. Best, Jacob
Dear Jacob, Thanks for your answer. I was Indeed referring to the ex-vivo drug sensitivity screening. This is what the auc relates to if I understood correctly? I was wondering about the rna seq data as in my mind I figured the drug concentration could impact the profile. Hope that makes sense as I am new to the topic. Best regards, Olivier
Hi Olivier, When you say drug administration, I'm not sure if you're referring to a patient's therapy or to the ex-vivo drug sensitivity screens. If you'll allow me to clarify: RNA seq, DNA seq, and drug sensitivity screenings are all done on the same biological sample. That sample is collected during the course of a patient's treatment, so a patient may-or-may-not be receiving a therapy at the time a sample is collected. Does that help? Best, Jacob
Hi Jacob, May I also ask about the rna seq data. Has it been donne prior to the drug administration or after? Thanks, Olivier
Hi Rasmus, > 1. Are all the clinical data obtained at the same time as the genomic and dose-response data, or are they possibly from diagnosis or inclusion? Most clinical data is obtained at the **same time as sample collection** (genomic and dose-response data are generated from the same sample). Note that some samples come from a common patient, so it is important that clinical data is updated at sample time, and not diagnosis or inclusion time. > 2. Are survival times used as response for sc2 given as time from diagnosis, inclusion or latest biosample (corresponding to genomic and dose response data)? Survival time is given as **time from diagnosis**. Note that some samples were collected after relapse, quite a long time after diagnosis. You can identify how long by looking at timeOfSampleCollectionRelativeToInclusion clinical_numerical field (since diagnosis usually happens soon after inclusion). > 3. Are all patients treated with the same regimen? **No.** We also don't provide the patients' regimen. This is a real shortcoming of our dataset - it will certainly be harder to predict outcome without knowing their treatment. Best, Jacob

More detailed information on data files page is loading…