Closed erthrall closed 1 year ago
I am not sure about the RIKEN FANTOM5 part.
For comparison, edgeR.design_matrix.csv includes design matrix used in edgeR, edgeR.csv includes all genes while edgeR.sig.csv includes significantly differential expressed genes only.
You may look at the code.
Hi erthrall, thanks for your question about RIKEN FANTOM5. Dr. Paula Hurley did the receptor-ligand analysis. I discussed with her and she said she described what she did in the Methods of our manuscript (PMID: 36229464):
"Receptor ligand interactions were inferred using receptor-ligands reported in RIKEN FANTOM5 database. Significantly increased DEGs in clusters 5, 6, and 11 (cancer cells) compared to cluster 12 (benign epithelial cells) were compared to ligands reported in RIKEN FANTOM5 database. The expression of corresponding receptors was assessed in cluster 20 cells."
If you need further clarifications, please feel free to email her: paula.hurley@vumc.org
Thanks, Hong Yuen
Thank you so much, Hong!
No problem Erthrall, with this I'll close this "Issue" thread. :^)
Hi again, I have a question about inferring receptor-ligand interaction with RIKEN FANTOM5. Could you give a brief description as to how you carried that out? There are a lot of atlases and tools, e.g. TET and USCS genome browser, available on the RIKEN FANTOM5 site, but it's also not very clear in the paper which data was used to carry out this part of the analysis (e.g. did you use CSV files of initial clusters with Seurat?).
In addition, could you please describe what the
edgeR.sig.csv
,edgeR.design_matrix.csv
,edgeR.DE.csv
, andedgeR.csv
do? I gather that these and the RNK files are the outputs of thein_clusters
andbetween_clusters
parts of the analysis, and the RNK files are analysed using GSEA.Thank you very much for reading!