phbradley/conga

Can CoNGA handle multiple samples?

lormmy opened this issue · 3 comments

I have 6 samples (6 csv and 6 h5), and I did not get logo.png using CoNGA. Can CoNGA handle multiple samples? What should I do?

Thanks for your question. I can see now it’s not documented in the README but yes it can!

Currently, this can be done in two easy steps using the setup_10x_clones.py and merge_samples.py scripts in the package. First, use setup_10x_clones.py to parse each of the filtered_contig_annotations.csv files into the conga clones format. Next, use merge_samples.py to generate a combined AnnData h5ad object and clones file.
It uses a tsvfile specifying the paths to the gex data, it’s companion clones file from step 1, and the gex data type of each sample. The merged object and clones file can then be run through the run_conga.py pipeline or run through analyses interactively.

https://github.com/phbradley/conga/blob/04f7b5dbdce4f81b841f35941f498e1dd55e5014/scripts/setup_10x_for_conga.py

https://github.com/phbradley/conga/blob/04f7b5dbdce4f81b841f35941f498e1dd55e5014/scripts/merge_samples.py

Thanks for your question. I can see now it’s not documented in the README but yes it can!

Currently, this can be done in two easy steps using the setup_10x_clones.py and merge_samples.py scripts in the package. First, use setup_10x_clones.py to parse each of the filtered_contig_annotations.csv files into the conga clones format. Next, use merge_samples.py to generate a combined AnnData h5ad object and clones file.
It uses a tsvfile specifying the paths to the gex data, it’s companion clones file from step 1, and the gex data type of each sample. The merged object and clones file can then be run through the run_conga.py pipeline or run through analyses interactively.

https://github.com/phbradley/conga/blob/04f7b5dbdce4f81b841f35941f498e1dd55e5014/scripts/setup_10x_for_conga.py

https://github.com/phbradley/conga/blob/04f7b5dbdce4f81b841f35941f498e1dd55e5014/scripts/merge_samples.py

Many thanks! It worked! I obtained the logo.png using the methods above.