CITE-seq analysis of immune cell subsets

Specificity	Fluorophore	Clone	Note
CD3	AF488	UCHT1	10uL of antibody used
CD8	APC-Cy7	SK1
CD4	AF700	RPA-T4
CD57	PE	QA17A04
CD56	APC	5.1H11
CD103	BV421	Ber-ACT8
Integrin 7	PE	FIB504
CD49a	APC	TS2/7	BioLegend
CD43	PE	CD43-10G7	BioLegend

Gating conditions for each of the validation experiments are shown in Figures 4D and 4E.
Post sorting, samples were each split into quintuplicates, and then cleaned up with 2x SPRI. Samples were then brought into reverse transcription in an adaptation of SMARTseq2 (Picelli et al., 2014 (link)) and SCRB-seq (Soumillon et al., 2014 (link)) as described here: https://dx.doi.org/10.17504/protocols.io.nkgdctw.
The pooled library was sequenced on an Illumina Nextseq (50 R1, 8 index, 34 R2). Post base calling, samples were aligned using a wrapper for DropSeqTools against the human reference hg19 to generate RNA counts matrices.
To assess the agreement between single-cell datasets and bulk-sorted experiments, we examined the top DE genes separating our gated populations in the CITE-seq reference dataset. We next visualized the relative expression of these genes in the heatmaps in Figures 4D and 4E. The bulk-sorted populations exhibited highly concordant relative expression patterns for DE genes as we observed in CITE-seq data.