-
Notifications
You must be signed in to change notification settings - Fork 30
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
liana_wrap function hangs #145
Comments
Hi @jjacob12, Hmmm, could it be that you are passing non-negative values? I see CST assay is used, perhaps make sure that the values are just log-normalized counts. Also, best to pass the name of the column containing the clusters for your cells via |
Ok, I'll try that out, just as soon I sort some compatibility issues between Seurat v4 and the latest tidyverse installation (required for OmnipathR, hence Liana to work) that mean Seurat plotting functions are failing. Will feedback soon. |
Hi @dbdimitrov,
The above code ran with no error messages and produced the expected output. After cell annotation the UMAP plot was as expected: Here is the structure of the Seurat object 'dyCbo.logNorm' used as input for
As per your advice I provided the argument Here's a picture of the metadata table which I used with 'celltype' labels: This is the output from running
|
Hi @jjacob12, yeah. I'm really not sure why this is the case. I assume it has to do with calculating the final p-values for CellPhoneDB, but I cannot reproduce this error... Can you try passing If it then hangs again, perhaps you could share a subset of your data, and I can then attempt to debug on my end. |
Hi @dbdimitrov, I have just tried:
This function threw an error quite quickly on trying to run it:
I attach the traceback which might help to troubleshoot: Please let me know if I need to share the relevant data - very happy to do so. |
Also, @dbdimitrov please send an email address to send the data to, if it's needed. |
Hi @jjacob12, daniel.dimitrov {a} uni-heidelberg.de |
Many thanks @dbdimitrov - I have shared a one-drive folder with the Seurat object which corresponds to the same object as 'dyCbo.logNorm.annotate' above. |
Hi @dbdimitrov I hope you were able to access the file. Please let me know if there was a problem with access. |
Hi @jjacob12, I was able to, thanks for sharing. I'm mid revisions of 2 papers but I will come back to you as soon as I find some time. Sorry for the delay. |
Sure @dbdimitrov. Good to know you accessed the file. |
Hi @jjacob12, Sorry it took me a while to come back to you. Using your subset data, even with 1000 permutations it worked on my laptop. Also, there is nothing obviously wrong with the data nor with your R session.
On my end, it also takes a bit of time once it reaches 100% but after a minute or so it returns the results as expected. I recommend using LIANA+ (extended liana in Python), and if you're interested I can share a couple of lines of code to convert a Seurat object into SingleCellExperiment and export it as an AnnData. Sorry, I couldn't be of more help.
|
Hi @dbdimitrov, that is really odd, but reassuring it works fine at your end with the data I sent you. "I recommend using LIANA+ (extended liana in Python), and if you're interested I can share a couple of lines of code to convert a Seurat object into SingleCellExperiment and export it as an AnnData." Yes please could you share this. Many thanks. |
Hi @jjacob12, I believe the easiest way is to use zellkonverter.
From there, it should be as easy as following the tutorials on the liana-py webpage, e.g.: Hope this helps. |
Tested your converted AnnData object in Python:
runs in ~ 30 sec :) |
Many thanks for this. |
Hi Daniel,Just to confirm Liana+ did the job. Thanks for your help and advice. Much appreciated.Dr John JacobSent from my iPhoneOn 6 Aug 2024, at 08:45, Daniel Dimitrov ***@***.***> wrote:
Closed #145 as completed.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Hi @jjacob12 great 😃 |
Hello,
I successfully installed Liana - thanks for your help with that! There's a new problem, which did not affect the above function when using the PBMC dataset in the Liana tutorial. Again I'm doing this on an HPC cluster with 1 core and I tried up to 25 GB RAM without success. One of my Seurat objects of interest is very heterogeneous with 3000+ cells and 15 annotated cell types.
Libraries I loaded:
This is the dataset of interest, which has previously been though SCtransform normalisation, then clustered:
Then I tried to run the
liana_wrap
function on the above dataset:It eventually got to 100% producing this message:
but even then continued to hang and I could see the ssh connection to the HPC cluster broke in the Terminal sessions I opened - example couple of lines below:
I ran this liana function multiple times to no avail after quitting and restarting a new RStudio session. Why is this happening and how do I fix it? Thanks again.
The text was updated successfully, but these errors were encountered: