Adding Salmon Alevin counts to clustered Seurat object
Hi Im trying to figure out an appropriate way to add Salmon Alevin count matrix to a previously created and clustered Seurat object, just to be able to see isoform distribution across clusters generated from scRNA-seq data. I can import Alevin counts and use those as the basis for Seurat clustering, but those results are quite different since its using all of the isoforms to do the clustering and differential calls. Any help greatly appreciated, thanks!
0 answers
No answers yet.
Log in to answer this question.
More posts like this
-
Spatial Transcriptomics: Analysis and differential expression across two slides
written by Vanish007 •
-
Can't reproduce harmony results in Seurat
written by fifty_fifty •
-
Differential Expression using Isoseq-supplemented reference transcriptome
written by Calum •
-
Seurat: pre-processing and differential markers
written by Rainer
-
Seurat subset cluster cell annotation
written by leranwangcs •
-
Going from Seurat object to DESeq2 analysis
written by learner-MD •
-
Trouble quantifying scRNAseq experiment
written by chien.chao.tw •
-
Analyzing digital gene expression data (DGE) from drop-seq pipeline with Seurat.
written by chilifan
-
Clustering Cells using known marker genes (Single cell RNAseq)
written by V
-
How to do unsupervised clustering using copy number variation data?
written by dr.chenway •
alevin returns gene level counts, not isoform-level, are you aware of that?
Ive been using the following approach:
https://tobitekath.github.io/DTUrtle/articles/Tabular_Muris_mouse_single-cell_preprocess.html
Basically converting Cellranger BAM files back to Fastqs and then mapping gencode.vM24.annotation.gtf transcript IDs. I believe in this approach Alevin provides transcript level counts, am I incorrect? Thanks.
Yes, I see, tricking the tx2gene map. Anyway, you need most likely a new Seurat object for this as it's not the same dimensions (rows) as a gene-level one. Just make a new one and transfer all the reduced dimensions and column-wise data to the new one.
Ok thanks for the tips. So I guess there is no way to include this within a full processed and clustered gene-level object? Seems the only way to do it otherwise is to start with a new object where the isoform level counts are the foundation of the clustering.