Read on Twitter
Running (EC)CITE-seq? #scRNAseq
Can I use FACS staining protocol? Should #CITEseq Abs be used at saturating conc? Can I reduce staining volume? Does cell# at staining matter? How can I minimize Background?
Tweetorial of https://www.biorxiv.org/content/10.1101/2020.06.15.153080v1
@SBKoralov @skinUCPH @NYGCtech
Can I use FACS staining protocol? Should #CITEseq Abs be used at saturating conc? Can I reduce staining volume? Does cell# at staining matter? How can I minimize Background?
Tweetorial of https://www.biorxiv.org/content/10.1101/2020.06.15.153080v1
@SBKoralov @skinUCPH @NYGCtech
When setting up #ECCITEseq we encountered many questions that we could not find answers to. Figuring we were not alone, we tested several conditions (tissue/conc./volume/cell #) and compiled the results into a manuscript that we hope will assist others! https://www.biorxiv.org/content/10.1101/2020.06.15.153080v1
We find that staining with high antibody conc. can cause unnecessarily high background signal and that concentrations of most antibodies can be reduced beyond general recommendations. Titration is best, but 0.6-1 µg/mL range appears to be a good starting point for most Abs tested
Unlike flow- and mass cytometry, oligo-conjugated antibodies rely on read counts by NGS. Thus, cost is determined by intensity and background signal. Many Abs can be reduced below saturation point without loss of information – reducing the sequencing requirements/cost.
Staining volume and cell density have highly limited effect on resulting antibody signals. Reducing staining volume only affects antibodies used at low concentrations targeting highly abundant epitopes and can be counteracted by reducing cell numbers at staining.
Background signal assayed from empty droplets accounts for a large % of total sequencing reads and is primarily derived from antibodies used at high concentrations. This implicates unbound antibodies as a major source of background warranting thorough washing protocols.
Finally, we benchmarked three workflows for counting antibody-derived tags. While all three pipelines yielded comparable read assignment, they showed drastic differences in their runtime. Kallisto-Bustools KITE ( @lpachter) is an order of magnitude faster than the other workflows.
The supplemental figures includes detailed data from each individual antibody at various conditions.
Data from the study available at FigShare: https://doi.org/10.6084/m9.figshare.c.5018987
All code used for analysis and to generate the figures: https://github.com/Terkild/CITE-seq_optimization
Data from the study available at FigShare: https://doi.org/10.6084/m9.figshare.c.5018987
All code used for analysis and to generate the figures: https://github.com/Terkild/CITE-seq_optimization
Questions, corrections, and suggestions to improve the manuscript are most welcome! Thank you for reading through to the end!
Great collaborative effort between @SBKoralov #KoralovLab, @skinUCPH, @NYGCtech ( @emimitou @psmibert) & @ThalesPapaG with (hopefully) more to come!
Great collaborative effort between @SBKoralov #KoralovLab, @skinUCPH, @NYGCtech ( @emimitou @psmibert) & @ThalesPapaG with (hopefully) more to come!
