Cell ratio between FACS and 10X

Question

[alexmascension]

Hi,

I want to run a single-cell RNA-seq experiment and I was wondering if there was any evidence about the number of cells that are filtered after FACS and the number of cells that are sequenced after library preparation, in this case with 10X.

I don't see any hallmark paper about these proportions for droplet methods, and the methods section from most of the papers don't even bother to include that information. This is very important for me because the number of cells that I have to isolate can vary a lot depending on this FACS/10X ratio, which can impact the experimental design of the study.

[GenoMax]

Please contact 10x technical support with this question since they are likely to have this information available.

Since cells differ in their sizes/stickyness/viability etc I am not sure if there is a generalizable inference that can be drawn with respect to FACS. I am not an experimental person so I am thinking out aloud.

You have probably seen general recommendations from 10x on cell numbers:
https://kb.10xgenomics.com/hc/en-us/articles/115001800523-What-is-the-minimum-number-of-cells-that-can-be-profiled-
https://kb.10xgenomics.com/hc/en-us/articles/360001378811-What-is-the-maximum-number-of-cells-that-can-be-profiled-

[igor]

Since cells differ in their sizes/stickyness/viability etc I am not sure if there is a generalizable inference that can be drawn

Yes, there are many factors involved. I've seen people load the same number of supposedly comparable cells and get 10,000 useable cells from one library and 1,000 from another one.

[ATpoint]

I would see whether you can get in contact with either a facility or a lab who has done 10x on either the same or a comparable cell type and hear what they have to say in terms of their experience. Our in-house experience with 10x (blood progenitors) was that these numbers that GenoMax linked were fairly accurate. It is crucial though to really count your cells with a counting chamber, the FACS counts are always an overestimation. Sort many more cells than actually necessary if you can, even if it is just for the purpose of counting several aliquots to minimize the counting error. Nothing beats having experience though, so really try and see whether you can get in contact with experienced folks using that cell type.

[igor]

get in contact with experienced folks using that cell type

Someone has to be the first to do it though.