Low expression of key transcripts in scRNAseq dataset after deeper sequencing
1
0
Entering edit mode
2.2 years ago
AW ▴ 10

I am having an issue with my single cell data set. This alert was shown after I requested additional reads per cell to be incorporated with the initial data we received. 10X output

We have previously performed single cell on this tissue so I am pretty certain that the issue is not due to trying to identify rare transcripts. We sent this data for sequencing to further test a hypothesis that a specific cell population is a major "expresser" of a gene of interest. We have shown this in multiple ways so far. I am seeing trends in the data that is in line with what we have previously found but the cell count is so low that my violin plots do not give a violin and I have sparse dots on the umap. See below. the goal is to show that genes 1,2 and 3 are expressed in the same cluster, cluster 0. Gene 3 is a cell-type identifier

the goal is to show that genes 1,2 and 3 are expressed in the same cluster, cluster 0. I am at my wits end with trying to represent this data. We have tried doing deeper sequencing to no avail. Does anyone have any idea is there is a way to represent this data as is in a way that makes sense. I don't think violin plots or umap plots will be useful here. Like I said, this is not our only piece of data so we are pretty confident that the finding is real.

Also, on the sequencing/10X side, is there anyway to improve data quality given this type of error? Note that we did not observe this error (picture 1) on the first round of sequencing.

genomics violin single-cell 10X plot • 735 views
ADD COMMENT
0
Entering edit mode
2.2 years ago

Are you saying you asked them to re-sequence the same libraries at a greater depth? I'm going to assume this here. Doing this you would expect the fraction of good barcodes (i.e. cells) to stay the same. Seems like contamination or something happened to the library. Either it got super degraded... or maybe there was barely any material left?

Otherwise if you performed the same whole 10x workflow on a new tissue sample it's more likely one of the wet-lab steps or GEM creation or extraction did not go as planned or even potentially sample you started with was really bad.

Unfortunately, if i was in this situation my plan would be to re-do the experiment :(

ADD COMMENT
0
Entering edit mode

Yes I asked for them to re-sequence the same libraries at a greater depth and incorporate the data with the original data. Thanks for your input.

ADD REPLY
0
Entering edit mode

Additional thought: What is so odd about this is that I collected and sequenced two samples on the same day. They were treated the same way with respect to tissue collection and one sample has high quality data and one does not. That is so weird, and it makes me wonder what exactly I should do differently if we decide to collect a new sample from this group and redo everything

ADD REPLY

Login before adding your answer.

Traffic: 1295 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6