Details
-
Type:
Task
-
Status: Closed (View Workflow)
-
Priority:
Major
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: None
-
Labels:None
-
Story Points:2
-
Epic Link:
-
Sprint:Summer 7
Description
In IGBF-3838, the PBMC dataset of 2700 single cells was obtained and viewed in IGB. Almost all of the single cell RNA sequences for the genes (mentioned in the ticket) did not line up with the reference gene models. The reads are misaligned from the boundaries of the exons in the gene model. The UMI count matrix data for the PBMC dataset was used in the Seurat Guided Clustering Tutorial . To investigate further, view the the reads of the the marker genes mentioned in the tutorial in IGB and check how many are aligned with the gene model.
Attachments
Issue Links
Activity
10XGenomics has their own browser called Loupe browser. The browser can not be used to visualize bam files, as far as I can tell, but just clustering visualizations. Looking into this browser might be useful. However, the input files to the browser is a cloupe file and I could not find one for the dataset we are looking at now. This may be set aside for another time.
The prospect of adding a Samtags filter to IGB was also discussed, or at least a cell barcode filter will help in verifying the UMI count matrix in IGB.
Images of the PBMC reads and respective depth graphs are added here: https://drive.google.com/drive/u/0/folders/1Bh2zG9UYoeFEGmYohk1c5-gFCRUE-zC4
Note: Depth graphs images have "DG" at the end of their filenames
The coverage of the ranges from 2.5k to 70k reads and 3' bias was observed on all marker genes. This dataset uses cellranger v1.1.0 and chemistry v1.0. Datasets created with newer cellranger and chemistry versions need to be found and observed (IGBF-3864). Bulk RNA seq should also be visualized for comparison. Moving this ticket to close.
I was counting the reads that align within the exon boundaries of the gene model and recording it in a spreadsheet. After talking to Dr. Nowlan Freese, I realized I have been counting only reads with no gaps and did not count the ones with some gaps. The counts needs to be updated with counts of reads that may have some gaps but are still within the exon boundaries.
We also discussed about the CellRanger pipeline and understanding how the UMI counts are counted for each feature (gene) for each cell or rather what reads are counted and what is discarded.