adugourd
commented
3 months ago Hi Guangyuan Liu,
Thank you very much for your interest in our tool.
I clarified the markdown of the cosmoR GitHub to make it easier to find what you are looking for.
For 1 and 2) you can find everything you are looking for here: https://github.com/saezlab/COSMOS_basic (input data and script that detail the pre-processing and where to obtain the raw source material)
For 3, the PKN is annotated for the signalling part using gene symbols. As long as you use HUGO gene symbols, there should be no mapping problem. Of course there will always be genes that are not part of our PKN, since it’s not exhaustive, but a good proportion should be there.
Cheers,
Aurelien
transform66
Dear Aurelien Dugourd, I hope this message finds you well. I am writing to seek your guidance on a few issues I've encountered while attempting to analyze my data using the COSMOS framework. As a fellow researcher utilizing this powerful tool, I'm eager to ensure that my analysis is conducted accurately and efficiently. Issue 1: Necessity of NCI60 Tutorial Preprocessing Firstly, I am unsure if I am obligated to follow the preprocessing steps outlined in the NCI60 tutorial(https://saezlab.github.io/cosmosR/articles/NCI60_tutorial.html) when preparing my data for COSMOS analysis. My concern stems from the specific nature of my proteomic dataset and whether it aligns seamlessly with the tutorial's guidelines. Could you please clarify if adhering strictly to the NCI60 tutorial preprocessing is a prerequisite or if there's flexibility in adapting the workflow to suit my unique dataset? Issue 2: Acquisition of cosmos_inputs.RData Assuming that the NCI60 tutorial preprocessing is indeed necessary, I've come across a reference to in the documentation, which seems to be a crucial input file. However, I'm unable to locate this file within the provided resources or the COSMOS GitHub repository. Could you kindly direct me to where I can access or if there's an alternative method to generate a similar input file tailored to my dataset? Issue 3: Gene ID Recognition in PKN Furthermore, if I opt not to follow the NCI60 tutorial, I've encountered a significant challenge where a substantial portion of my proteomic dataset's gene IDs are not recognized by PKN, thereby hindering the smooth execution of COSMOS. To address this issue, could you suggest any strategies or tools that might assist in converting or annotating my gene IDs in a manner that enhances their recognition by PKN? Alternatively, are there any known limitations or compatibility issues with specific types of gene IDs that I should be aware of? Your insights and recommendations would be invaluable in overcoming these obstacles and ensuring the success of my COSMOS analysis. Please let me know if there's any additional information I can provide to facilitate your assistance. Thank you very much for your time and expertise. I look forward to your response.