VCCRI / Sierra

Discover differential transcript usage from polyA-captured single cell RNA-seq data
GNU General Public License v3.0
49 stars 17 forks source link

umap coordinates for NewPeakSeurat #43

Open leeanapeters opened 3 years ago

leeanapeters commented 3 years ago

Hi there, thank you for creating this interesting package!

I am attempting to create my peak level Seurat object. I initially received an error when using PeakSeuratFromTransfer that there were no matching barcodes despite arranging them in the same order when aggregating peaks (the combined prepared Seurat object is has barcode suffixes _1-1, _2-1 etc, which was how I understood the Sierra barcode appending worked as well, but maybe I am incorrect?)

When I tried NewPeakSeurat I encountered another error. It seems to want tsne coordinates even though I am supplying umap coordinates in the function.

Here is what I did along with the error

umap.coords <- query[["proj.umap"]]@cell.embeddings

peaks.seurat <- NewPeakSeurat(peak.data = peak.counts,

  • annot.info = peak.annotations,
  • cell.idents = populations,
  • umap.coords = umap.coords,
  • min.cells = 0, min.peaks = 0) [1] "Creating Seurat object with 35146 peaks and 60227 cells" Warning: The following arguments are not used: row.names [1] "Preparing feature table for DEXSeq" Performing log-normalization 0% 10 20 30 40 50 60 70 80 90 100% [----|----|----|----|----|----|----|----|----|----| **| [1] "No t-SNE coodinates included" Error in umap.coords[colnames(peaks.seurat), ] : subscript out of bounds

Any help would be appreciated!

Thanks

leeanapeters commented 3 years ago

After digging a bit more, I am going to try and re-run the aggregatepeaks portion and provide manual barcode labels now, hoping that solves the issue.

pratarora commented 3 years ago

Hi @leeanapeters

I think this issue was resolved in the new version available from github. Try installing the newer version. Also check #40 .

rj-patrick commented 3 years ago

Hi @leeanapeters,

For aggregating multiple runs, by default with barcodes will be appended with '-1', '-2' etc. Looks like this is not the case for your data, so you will need to update the suffixes manually.

Regarding the t-SNE coordinates, yes this should be fixed in the current version as noted by @pratarora.

Cheers, Ralph