Reference

Population trees

One hierarchy per species: the populations people actually report, in the order they are gated, with the marker logic that defines each one. These are starting points, not answers. The panel in front of you decides which nodes are reachable.

Species	Tree	Populations	Distinct stains	Source
Human	Human PBMC / whole blood (HIPC consensus)	36	21	verified
Mouse	Mouse spleen / lymph node	37	34	unverified
Macaque	Macaque PBMC (rhesus / cynomolgus)	21	16	unverified
Rat	Rat blood / spleen	19	15	unverified
Dog	Canine PBMC (OMIP-065)	20	14	verified
Dog (CD3 pan-T)	Canine PBMC (OMIP-065, CD3 pan-T variant)	20	13	unverified

Human Human PBMC / whole blood (HIPC consensus)

The population set the standardisation consortia actually report, with the marker logic taken verbatim from Table 2 of Finak et al. 2016 (the Human ImmunoPhenotyping Consortium cross-site validation) and, where HIPC does not resolve a population, from the sources named on the node. Two things this tree deliberately does not smooth over. The consortia do NOT agree on a single pre-gating order. Finak describes it as "ungated -> lymphocytes/singlets -> live -> lineage -> activation/memory", but the exact order of singlet, viability and scatter gates varies paper to paper. That variation is itself a measured source of cross-site disagreement, so the order below is the common one, not a standard. The finding that matters from Finak 2016: cross-site variability fell substantially under centralised analysis, i.e. **gating, not staining, is the dominant source of variance between labs**. Standardising reagents without standardising the hierarchy does not reproduce.

Verified against source

Finak G, et al. Standardizing Flow Cytometry Immunophenotyping Analysis from the Human ImmunoPhenotyping Consortium. Sci Rep. 2016;6:20686. Panels designed in Maecker HT, McCoy JP, Nussenblatt R. Standardizing immunophenotyping for the Human Immunology Project. Nat Rev Immunol. 2012;12:191-200 (see the erratum, nri3229).

Source · PMID 26861911 · Table 2 (population definitions); Table 1 (the five 8-colour panels)

Monocyte subsets follow Ziegler-Heitbrock et al., Blood 2010 (PMID 20628149), the nomenclature authority - HIPC's own DC/mono/NK table reports only a bare CD14/CD16 quadrant grid and does not resolve all three subsets. The HIPC Th1/Th2/Th17 panel is published but was dropped from the 2016 validation for excessive variability; nothing from it is included here.

All events —
Single cells FSC-H FSC-A
Often followed by SSC-A vs SSC-H. Order relative to the live gate is not standardised.
Live Viability-
Every HIPC panel puts the viability dye in the FITC channel.
Lymphocytes FSC-A SSC-A
T cells CD3+
B cells CD3- CD19+
CD20 is the usual second pan-B marker; HIPC gates CD19+ then CD20+ for the subsets.
NK cells CD3- CD56+
CD16 is also on non-classical monocytes and neutrophils, so CD14 must be excluded before calling CD16+ cells NK.
Monocytes CD3- CD19- CD56- CD14+
Lineage-dump composition is NOT standardised between consortia and directly changes monocyte and DC frequencies.
Dendritic cells CD3- CD19- CD14- HLA-DR+
CD4+ T cells CD4+ CD8-
Drawn on the CD4 vs CD8 plot, not as a box on each axis. APC spreads into AF700 on this panel, so a CD8 upper bound cuts off the CD4 cells whose brightness leaks across.
CD8+ T cells CD8+ CD4-
CD4 naive CCR7+ CD45RA+
CD4 central memory CCR7+ CD45RA-
CD4 effector memory CCR7- CD45RA-
CD4 effector CCR7- CD45RA+
HIPC calls this gate 'Effector'; most of the field calls it TEMRA. Same gate, different label - do not silently map the names onto each other. EuroFlow's PID panel splits T memory on CD62L/CD45RO instead, which is parallel but not identical.
CD8 naive CCR7+ CD45RA+
CD8 central memory CCR7+ CD45RA-
CD8 effector memory CCR7- CD45RA-
CD8 effector CCR7- CD45RA+
Activated CD4 T cells CD38+ HLA-DR+
Activated CD8 T cells CD38+ HLA-DR+
Regulatory T cells CD127lo CD25hi CCR4+
HIPC's Total Treg requires CCR4 on top of CD127-lo/CD25-hi; the ONE Study uses CD25++CD127-lo without CCR4. NOTE: neither consortium panel includes FoxP3 - they are deliberately surface-only so whole blood can be counted without fix/perm. FoxP3 is the lineage-defining transcription factor and belongs in a research panel, but it is not what these citations mean by Treg. Gate placement on the CD127/CD25 continuum is the single largest reproducibility problem in Treg quantification.
CD20+ B cells CD20+
Naive B cells CD27- IgD+
Non-switched memory B cells CD27+ IgD+
Class-switched memory B cells CD27+ IgD-
Double-negative B cells CD27- IgD-
Transitional B cells CD24hi CD38hi
Transitional and plasmablasts share CD24-hi CD38-hi and are separated by CD20 - transitional are CD20+, plasmablasts CD20-. Easy to get wrong.
Plasmablasts CD20- CD24hi CD38hi
Classical monocytes CD14hi CD16-
Ziegler-Heitbrock nomenclature (CD14++CD16-). The classical/intermediate boundary is a continuum and gate placement is contested.
Intermediate monocytes CD14hi CD16+
Non-classical monocytes CD14+ CD16hi
CD56dim NK cells CD56dim CD16+
Cytotoxic; about 90% of blood NK.
CD56bright NK cells CD56bright CD16lo
Cytokine-producing; about 10% of blood NK.
Myeloid dendritic cells CD11c+ CD123-
Plasmacytoid dendritic cells CD11c- CD123+
The ONE Study adds BDCA2/CD303 to confirm pDC and BDCA3/CD141 for cDC1.

Mouse Mouse spleen / lymph node

The standard murine lymphoid hierarchy. Assembled from the primary literature cited per node rather than transcribed from one figure, so `verified` is false - check a node against its citation before leaning on it. Three things that bite, encoded as nodes rather than left in a footnote: Fc block (anti-CD16/32, clone 2.4G2) before any myeloid staining. Not optional. B220 alone is not a B cell gate. It is also on plasmacytoid DCs, activated T cells, NK progenitors and the double-negative T cells of lpr/lpr mice. Pair it with CD19 - the pDC confounder is specifically CD19-negative. Gr-1 (clone RB6-8C5) binds both Ly6G and Ly6C and is not a neutrophil marker. Use 1A8 for Ly6G.

Not verified — check before you lean on it

Assembled from: Hardy RR et al., J Exp Med 1991 (BM B-cell fractions); Wrammert J et al., Eur J Immunol 2002;32:97-103 (Ly6C on plasma cells); Nakano H, Yanagita M, Gunn MD, J Exp Med 2001;194:1171-8 (pDC); Daley JM et al., J Leukoc Biol 2008 (1A8 vs RB6-8C5); Chiossone L et al., Blood 2009 (NK maturation); Guilliams M et al., Nat Rev Immunol 2014 (DC nomenclature). Published panels to cite instead where they fit: OMIP-095 (40-colour, all major leukocyte populations in murine lymphoid tissues) and OMIP-076 (T, B and antibody-secreting cell subsets).

Source

Not transcribed from a single figure. Each node names its own source; OMIP-095 is the closest thing to a single citable whole-hierarchy reference and should be read before this file is treated as settled.

All events —
Stable acquisition —
Cells FSC-A SSC-A
Excludes debris and RBC ghosts. Bone marrow additionally needs a TER-119 dump for erythroid cells.
Single cells FSC-H FSC-A
Live Viability-
Amine-reactive fixable dye, or 7-AAD/DAPI if unfixed.
Leukocytes CD45+
Clone 30-F11. Near-redundant in spleen and LN (>95% CD45+); load-bearing in tissue digests and tumours. For tissue-resident vs vascular, i.v. anti-CD45 3-5 min before sacrifice labels the intravascular pool - parenchymal cells are i.v.-CD45-negative.
T cells CD3e+
Clone 145-2C11 or 17A2.
alpha-beta T cells TCRb+
gamma-delta T cells TCRgd+
Clone GL3 or UC7-13D5. See OMIP-057 for a dedicated 14-colour development panel.
CD4+ T cells CD4+
CD8+ T cells CD8a+
Naive T cells CD44lo CD62Lhi
Mouse uses CD44/CD62L, not the human CCR7/CD45RA pair.
Central memory T cells CD44hi CD62Lhi
Effector memory T cells CD44hi CD62Llo CD127+
CD127-negative (often KLRG1+) in the same gate are short-lived effectors, not memory.
Regulatory T cells CD25+ FoxP3+
CD25 is transiently induced on activated conventional T cells, so CD25 alone over-calls Tregs in inflamed tissue. FoxP3 (FJK-16s) is intranuclear - fix/perm, and no viable sorting. Human-style CD127-low gating is much less used in mouse. Helios and Nrp1 as thymic- vs peripherally-derived discriminators are genuinely contested.
B cells CD19+ CD45R+
B220/CD45R (RA3-6B2) AND CD19 (1D3 or 6D5). Use both - see the tree description.
Transitional B cells CD93+
T1 IgM-hi CD23- CD21-; T2 IgM-hi CD23+ CD21-int; T3 IgM-lo CD23+.
Mature B cells CD93-
Follicular B cells CD23+ CD21lo
IgD-hi IgM-lo.
Marginal zone B cells CD21hi CD23lo
IgM-hi IgD-lo.
B-1 cells CD19hi CD45Rlo CD43+
Peritoneal and pleural cavity; rare in spleen. IgM-hi IgD-lo CD23-. No established Ly6C positivity on B-1 cells - do not infer it.
B-1a cells CD5+
B-1b cells CD5-
Plasmablasts / plasma cells CD138hi CD45Rlo Ly6C+
THE B220-lo, Ly-6C+ POPULATION. Ly6C is absent from resting B cells, is induced by LPS (blocked by anti-CD40 or anti-Ig, so it tracks T-independent/extrafollicular differentiation) and is present on ex vivo plasma cells from spleen, bone marrow and lamina propria - Wrammert et al., Eur J Immunol 2002;32:97-103, PMID 11754008. Plasma cells independently downregulate B220 as they gain CD138. CAUTION: CD138-hi/B220-lo alone is not sufficient - that gate in bone marrow holds a substantial Blimp1-negative fraction. Add TACI, CD98, intracellular Ig or a Blimp1 reporter.
Age-associated B cells CD21- CD23- CD11c+ CD11b+
B220-POSITIVE, not B220-lo - so ABCs are NOT the B220-lo/Ly6C+ population; that is plasmablasts. No literature establishes Ly6C as an ABC marker. T-bet is usually included but is contested: B-cell-intrinsic Tbx21 deletion did not prevent ABC generation and the knockouts were phenotypically identical to wild type (Du et al., Eur J Immunol 2019; Nickerson et al., J Exp Med 2023).
NK cells CD3e- NKp46+
NKp46/NCR1 (29A1.4) is the pan-strain choice. NK1.1 (PK136) is STRAIN-DEPENDENT and is an antibody-epitope problem, not receptor absence: reactive in C57BL/6, C57BL/10, FVB/N, NZB; NOT reactive in BALB/c, AKR, CBA/J, C3H, DBA/1, DBA/2, NOD, 129. There is no functional mouse CD56 gate.
Mature NK cells CD27- CD11b+
Four-stage CD27/CD11b maturation: DN -> CD27+CD11b- -> CD27+CD11b+ -> CD27-CD11b+ (most mature). Chiossone et al., Blood 2009.
Myeloid cells CD11b+
Neutrophils Ly6G+
Clone 1A8, which is Ly6G-specific. Gr-1 (RB6-8C5) binds Ly6G AND Ly6C and in vivo depletes neutrophils plus Ly6C-hi monocytes.
Classical monocytes Ly6G- Ly6Chi
CCR2-hi CX3CR1-lo. The mouse equivalent of human CD14++CD16- classical monocytes; the correspondence is transcriptionally supported but imperfect.
Non-classical monocytes Ly6G- Ly6Clo
CX3CR1-hi CCR2-lo, Nr4a1-dependent.
Eosinophils SiglecF+ Ly6G-
SSC-hi. In lung, SiglecF is ALSO high on alveolar macrophages - separate on CD11c and autofluorescence.
Macrophages CD64+ F4_80+
F4/80 is strongly tissue-dependent: maximal on Kupffer, splenic red pulp and peritoneal macrophages, but LOW OR NEGATIVE on alveolar, marginal zone and T-zone macrophages, and present on gut and peritoneal eosinophils. Use CD64/MerTK for identity and SiglecF/CCR3 to exclude eosinophils.
Conventional dendritic cells CD11c+ MHCII+ CD26+ CD64-
cDC1 XCR1+
Also CD8a+ / CD103+ / CLEC9A+ depending on tissue.
cDC2 CD11b+ CD172a+
Plasmacytoid dendritic cells CD11cint CD45R+ Ly6Chi SiglecH+ CD19-
THE B220+Ly6C+ CONFOUNDER. pDCs are B220-positive and Ly6C-positive but CD19-NEGATIVE, which is the cleanest way to keep them out of a B cell gate. PDCA-1/BST2 alone does not fix it - PDCA-1 also marks B-cell-biased lymphoid progenitors and pre-pro-B cells, so it must be used together with Ly6C and CD11c.

Macaque Macaque PBMC (rhesus / cynomolgus)

The macaque hierarchy, which looks like the human one and is wrong in three specific places. Anyone porting a human panel to NHP hits all three. CD56 is not an NK marker in macaque - it marks MONOCYTES. Rhesus NK cells are mostly CD56-NEGATIVE. The NK definition is CD3-negative, CD8-alpha-positive, NKG2A/C-positive (clone Z199), and the dump has to exclude CD20 and CD14 explicitly, not just CD3. Because virtually all macaque NK cells are CD8-alpha-positive, a plain "CD8" gate contaminates the CD8 T cell gate with NK cells. Use CD8-beta (2ST8.5H7) for T cells. The alpha/beta distinction matters far more here than in human. Naive/memory runs on CD28 vs CD95, not CCR7/CD45RA. CD45RO is species- and even CONJUGATE-dependent in macaque: the FITC conjugate of UCHL1 fails on macaque cells while PE/APC/PerCP-Cy5.5 conjugates of the same clone work. Conjugate is an independent variable here, not a footnote. Clone choice is not interchangeable. CD3 must be SP34 or SP34-2 - UCHT1 and OKT3 are not reliable substitutes, and this is the classic NHP pitfall. Note SP34 is mitogenic, so avoid it if the stained cells are going back into culture.

Not verified — check before you lean on it

Hierarchy follows OMIP-028 (Activation panel for rhesus macaque NK cell subsets, Cytometry A 2015;87(10):890-3, PMID 26218174) and the macaque OMIP series (005, 016, 026, 035, 052, 075). Memory definition: Pitcher CJ et al. Development and homeostasis of T cell memory in rhesus macaque. J Immunol 2002;168(1):29-43, PMID 11751943. Cross-reactivity caveats: Bjornson-Hooper ZB et al., Front Immunol 2022;13:867015.

Source · PMID 11751943

Assembled from reported readings, not transcribed from a figure I opened. The NIH Nonhuman Primate Reagent Resource reactivity database is the usual clone reference, but its own disclaimer says the data were contributed by investigators or manufacturers and are not confirmed by them - and it records WHETHER a clone stains, not WHAT it stains. Read 'cross-reactive: yes' as 'binds something'.

All events —
Stable acquisition —
Single cells FSC-H FSC-A
Live Viability-
Leukocytes CD45+
Clone D058-1283, raised against rhesus; reactive with rhesus, cynomolgus, African green monkey and baboon. It does NOT stain human cells, so it is not a shared human/NHP backbone - running human and macaque samples side by side needs two different CD45 clones.
Lymphocytes FSC-A SSC-A
T cells CD3+
SP34 or SP34-2 only. Raised against human CD3-epsilon but binds a macaque-conserved epitope; UCHT1 and OKT3 are not reliable substitutes.
CD4+ T cells CD4+
Clone L200 is what the published macaque OMIPs use. OKT4 and SK3 are also reported reactive; RPA-T4 is reported non-reactive in rhesus. These calls trace to vendor datasheets rather than head-to-head benchmarking.
CD8+ T cells CD8b+
CD8-BETA (clone 2ST8.5H7) deliberately, not CD8-alpha. Macaque NK cells are CD8-alpha-positive, so an SK1/RPA-T8 gate pulls NK cells into the CD8 T gate.
Naive T cells CD28+ CD95-
Pitcher et al. 2002. CD95-lo CD28-hi.
Central memory T cells CD28+ CD95+
Effector memory T cells CD28- CD95+
Enriched at extralymphoid effector sites, holds most CMV-specific cells, rises with age.
CCR7+ memory refinement CCR7+
Clone G043H7. Use CD28/CD95 to call naive vs memory FIRST, then refine with CCR7 and optionally CD45RA - not the other way round. CD45RA (5H9, HI100) cross-reacts but also stains monocytes and some granulocytes, and CD45RA+ lymphocyte frequency is higher than in human, so the naive/memory split does not sit where a human panel expects.
B cells CD3- CD20+
CD20 (clone 2H7). Most anti-human CD19 clones do NOT cross-react - this is community consensus with thin formal citation, i.e. the evidence is that nobody uses CD19 in macaque rather than a published failure report. Note CD2 is present on macaque B cells, unlike human.
Monocytes CD3- CD14+
Clone M5E2. CD56 also marks macaque monocytes - the opposite of its human meaning. CD33, which marks monocytes and cDC in human, prominently marks GRANULOCYTES in macaque.
Dendritic cells CD3- CD14- HLA-DR+
HLA-DR clone L243 or LN3.
Myeloid dendritic cells CD11c+ CD123-
CD11c clone 3.9.
Plasmacytoid dendritic cells CD11c- CD123+
CD123 clone 6H6.
NK cells CD3- CD20- CD14- CD8a+ NKG2A+
THE POPULATION A HUMAN PANEL GETS WRONG. Defined by CD8-alpha-alpha and NKG2A/C (clone Z199) on CD3-negative lymphocytes, per OMIP-028. The dump excludes CD20 and CD14 explicitly, not just CD3. Virtually all rhesus peripheral NK are CD8-alpha+ and MOST ARE CD56-NEGATIVE. NKp46 (BAB281) is NOT reliable in macaque - not all peripheral rhesus NK express it, the opposite of its status in mouse.
CD16+ NK cells CD16+
Clone 3G8. CD16 and CD56 are used DOWNSTREAM to split subsets, never to define the NK gate itself.
CD16- NK cells CD16-
Enriched in mucosa and lymphoid tissue relative to blood.

Rat Rat blood / spleen

Rat is genuinely reagent-limited and has almost no standardisation behind it. There are only two rat OMIPs in the whole series - OMIP-041 (microglia) and OMIP-092 (bronchoalveolar lavage) - and neither is a general peripheral hierarchy. What follows is the de facto backbone from Barnett-Vanes et al. 2016, which is the closest thing rat has to a consensus panel, not a standard. Most rat markers are still named by hybridoma clone rather than CD number (Oxford OX-series, Amsterdam ED-series, Groningen HIS-series), so the clone is recorded on each node - it is load-bearing here in a way it is not for human. Three traps that make a mouse- or human-shaped tree wrong in rat: Rat CD4 is robustly expressed on monocytes - the CD43-hi subset is uniformly CD4-positive - so CD4 without CD3 does not define a helper T cell. Rat monocytes are largely CD161-positive (68-98% in blood). CD161 alone does not identify NK cells; CD3, CD45R and His48 all have to be excluded first. Rat monocytes carry no CD14, no CD16 and no Ly6C. The subset axis is CD43 against His48, and the non-classical subset predominates (~93%) - the inverse of human, and unlike the roughly even mouse split.

Not verified — check before you lean on it

Barnett-Vanes A, Sharrock A, Birrell MA, Rankin S. A Single 9-Colour Flow Cytometric Method to Characterise Major Leukocyte Populations in the Rat: Whole Blood, Spleen and Bone Marrow. PLoS ONE 2016;11(1):e0142520. Monocyte subset biology cross-checked against Hume DA et al., J Immunol 2021 (PMID 33965905). CD45RC memory axis: Powrie F & Mason D (PMID 2784479); Bunce C & Bell EB, J Exp Med 1997.

Source · PMID 26764577

Assembled from a reported reading of the source, not transcribed from a figure I opened. Verify against the paper before relying on it. Specific known hazard: at least one secondary summary REVERSES the rat classical/non-classical monocyte labels - if precision matters, report the raw CD43/His48 phenotype and skip the classical/non-classical naming entirely.

All events —
Cells FSC-A SSC-A
Single cells FSC-H FSC-A
Live Viability-
Leukocytes CD45+
Clone OX-1.
T cells CD3+
No single dominant clone. 1F4 and G4.18 bind DIFFERENT epitopes - 1F4 pre-incubation blocks G4.18 - so they are not interchangeable. REA223 is the recombinant option. TCR alpha-beta is R73 (~97% of peripheral T cells, no gamma-delta reactivity); TCR gamma-delta is V65.
CD4+ T cells CD4+
Clone W3/25 or OX-35. Only a helper gate because CD3 came first - rat CD4 is also on monocytes.
CD8+ T cells CD8a+
Clone OX-8, which is also on most rat NK cells and on CD8-alpha+ DC subsets. CD8-beta (clone 341) is the T-restricted alternative but is much less stocked.
Naive CD4 T cells CD45RChi
Rat uses CD45RC (MRC OX-22), NOT the mouse CD44/CD62L pair or the human CCR7/CD45RA pair. CD45RC is also on B cells and most CD8 T cells, so gate CD4 first.
Memory CD4 T cells CD45RClo
Regulatory T cells CD25+ FoxP3+
CD25 is clone OX-39, FoxP3 is FJK-16s. Contested: in Wistar rats only about 60% of CD4+FoxP3+ cells co-express CD25, so a CD25-gated Treg count is an undercount. Human-style CD127-low gating does NOT transfer reliably to rat.
CD8+ regulatory T cells CD45RClo
A distinctive rat compartment with no direct mouse analogue. Included because it is a real population people report, not because it is easy to gate.
B cells CD3- CD45R+ CD43-
CD45R here is clone HIS24 - B cells from early bone marrow development through the periphery, but NOT plasma cells. CD45RA (MRC OX-33) is a CD45 isoform found only on rat B lymphocytes and is the cleanest single rat B marker if you have it. Anti-rat CD19 is effectively unavailable as a working flow reagent - catalogue hits are almost all rat-anti-MOUSE CD19, which is the reverse of what is wanted. Treat 'rat CD19' in a panel as a red flag. Outside blood and spleen, CD45R+CD11b+ cells exist that are not B cells.
NK cells CD3- CD161+ His48-
The His48 exclusion is not optional: rat monocytes are 68-98% CD161-positive, so CD3-/CD161+ alone is contaminated. Clones 10/78 and 3.2.3 are labelled NKR-P1A-specific but both stain NKR-P1A and NKR-P1B equally, so read 'CD161a+' in rat papers as 'NKR-P1A/B+'. Most rat NK are CD8-alpha+. NKp46 (clone WEN23) is the only truly NK-specific rat marker but commercial availability is poor.
Neutrophils CD43lo His48hi
Separated from classical monocytes by SSC-hi, not by a marker. RP-1 is genuinely neutrophil-specific (no monocytes, macrophages or eosinophils) if you can add it. There is no rat Ly6G.
Classical monocytes CD43lo His48hi CD4-
The Ly6C-hi-like subset; about 98% CD161+. His48 on rat monocytes is the functional analogue of Ly6C on mouse monocytes. SEE THE SOURCE NOTE - one secondary summary reverses these labels.
Non-classical monocytes CD43hi His48lo CD4+
Uniformly CD4+, CD11c+, MHC-II+, CX3CR1-hi CCR2-lo. Predominates in rat at about 93% of monocytes - the inverse of human.
Tissue macrophages CD68+
ED1 = CD68. ED2 = CD163 marks resident tissue macrophages and is ABSENT from monocytes, alveolar macrophages and microglia. ED3 = CD169/sialoadhesin.
Dendritic cells CD103+ MHCII+
OX-62 is the usual rat DC reagent (CD103) but is also on gamma-delta T cells, so it needs a CD3 exclusion. MHC-II in rat is RT1B (OX-6, the I-A homologue, widely available) and RT1D (OX-17, the I-E homologue) - these are different molecules and OX-17 should not simply be labelled 'MHC-II'. CD11b is WT.5 and CD11c is 8A2; OX-42 recognises a shared CD11b/CD11c epitope and CANNOT discriminate them.

Dog Canine PBMC (OMIP-065)

Transcribed from Figure 1 of OMIP-065, panel A-I, which is a published and validated 14-colour gating strategy on cryopreserved canine PBMC. The ordering matters and is not arbitrary. CD14+ monocytes are removed FIRST, before any lymphocyte gate, because canine CD4 is not a T-helper marker on its own - it sits on monocytes and neutrophils too. A human-shaped tree that goes straight to CD4 vs CD8 produces a "T helper" gate full of myeloid cells.

Verified against source

Pantelyushin S, Ranninger E, Bettschart-Wolfensberger R, vom Berg J. OMIP-065: Dog Immunophenotyping and T-Cell Activity Evaluation with a 14-Color Panel. Cytometry A. 2020;97(10):1024-1027.

All events —
Stable acquisition —
Time vs a bright channel. Cuts fluidics disturbances at start/end of the tube.
Leukocytes FSC-A SSC-A
Scatter gate excluding debris. 86.2% in the paper's example.
Single cells FSC-H FSC-A
FSC-H vs FSC-A. Doublets carry two cells' fluorescence and invent double-positives.
Live Viability-
Aqua viability dye. Dead cells bind antibody non-specifically in every channel.
Monocytes CD45+ CD14+
Gated out before lymphocytes, deliberately - see the tree description.
CD45+ CD14- leukocytes CD45+ CD14-
The non-monocyte leukocyte pool that every lymphoid gate below hangs off.
MHCII+ monocytes MHCII+ CD4-
Fig 1C. Canine monocytes are primarily MHCII+CD4-, with smaller MHCII+CD4+ and MHCII-CD4+ fractions.
MHCII+ CD4+ monocytes MHCII+ CD4+
CD4 on a monocyte, not a T cell. The reason CD14 is gated first.
B cells CD22+
CD22 clone RFB-4 in the published panel. CD21 is the common alternative canine B marker; they are not interchangeable.
T cells CD5+
CD5 is the pan-T gate here, not CD3. CD3 is downregulated after activation and separates poorly post-stimulation; CD5 is not. Low CD5 is present on canine NK cells.
CD5- CD22- cells CD5- CD22-
Double-negative pool: mostly APCs plus neutrophils. Fig 1E splits it.
Antigen-presenting cells MHCII+
Fig 1E. 25.9% of the double-negative pool in the example.
Neutrophils MHCII- CD4+
Canine neutrophils are CD4+. In humans this gate would read as T helper cells.
NK cells GranzymeB+ CD3-
Fig 1F. Putative NK - the paper is explicit that no publicly agreed canine NK marker exists, so this is a working definition, not a settled one. Some canine NK are CD8a+ and CD5-low.
CD4+ T cells CD4+
Only trustworthy as T helper because CD14+ monocytes and CD5- neutrophils were already removed above.
CD8+ T cells CD8a+
Regulatory T cells FoxP3+ CD25+
Fig 1H. Intracellular FoxP3 - needs fix/perm, and the FMO must go through the same fix/perm as the full stain.
Eomes+ CD8 T cells Eomes+
Fig 1I. Many canine CD8+ T cells express Eomes.
Proliferating CD8 T cells Ki67+
Fig 1I. A minority are Ki-67+ at rest; this is the proliferation readout.

Dog (CD3 pan-T) Canine PBMC (OMIP-065, CD3 pan-T variant)

Not verified — check before you lean on it

Pantelyushin S, Ranninger E, Bettschart-Wolfensberger R, vom Berg J. OMIP-065: Dog Immunophenotyping and T-Cell Activity Evaluation with a 14-Color Panel. Cytometry A. 2020;97(10):1024-1027.

All events —
Stable acquisition —
Time vs a bright channel. Cuts fluidics disturbances at start/end of the tube.
Leukocytes FSC-A SSC-A
Scatter gate excluding debris. 86.2% in the paper's example.
Single cells FSC-H FSC-A
FSC-H vs FSC-A. Doublets carry two cells' fluorescence and invent double-positives.
Live Viability-
Aqua viability dye. Dead cells bind antibody non-specifically in every channel.
Monocytes CD45+ CD14+
Gated out before lymphocytes, deliberately - see the tree description.
CD45+ CD14- leukocytes CD45+ CD14-
The non-monocyte leukocyte pool that every lymphoid gate below hangs off.
MHCII+ monocytes MHCII+ CD4-
Fig 1C. Canine monocytes are primarily MHCII+CD4-, with smaller MHCII+CD4+ and MHCII-CD4+ fractions.
MHCII+ CD4+ monocytes MHCII+ CD4+
CD4 on a monocyte, not a T cell. The reason CD14 is gated first.
B cells CD22+
CD22 clone RFB-4 in the published panel. CD21 is the common alternative canine B marker; they are not interchangeable.
T cells CD3+
VARIANT. OMIP-065 and dog.toml gate pan-T on CD5, for a real reason: CD3 is downregulated after activation and separates poorly post-stimulation, while CD5 is not, and canine NK cells carry low CD5. This file overrides that because in THIS panel the CD5 conjugate (PerCP-eFluor710) is dim and the CD3 conjugate (AF750) is bright - on the 26-JUL Normal PBMC control, CD3 marks 85.2% of CD45+ leukocytes and CD5 only 21.1%, a four-fold undercount on material that stained well. Reagent brightness, not canine biology. THE CONFOUND IS REAL AND UNRESOLVED: CD3 in the Vax2 timecourse falls 41,948 -> 13,712 -> 17,085 (99th percentile) from pre to 2h to 6h, which is exactly what activation-induced CD3 downregulation looks like, so a CD3 gate may undercount activated T cells in the post-dose samples this study is about. Read both trees and compare; neither is right on its own here.
CD3- CD22- cells CD3- CD22-
Double-negative pool: mostly APCs plus neutrophils. Fig 1E splits it.
Antigen-presenting cells MHCII+
Fig 1E. 25.9% of the double-negative pool in the example.
Neutrophils MHCII- CD4+
Canine neutrophils are CD4+. In humans this gate would read as T helper cells.
NK cells GranzymeB+ CD3-
Fig 1F. Putative NK - the paper is explicit that no publicly agreed canine NK marker exists, so this is a working definition, not a settled one. Some canine NK are CD8a+ and CD5-low. VARIANT NOTE: here CD3- is also the T cell gate's complement, so NK and the CD3- CD22- bucket overlap by construction; in dog.toml they are independent because T cells are gated on CD5 there.
CD4+ T cells CD4+
Only trustworthy as T helper because CD14+ monocytes and CD5- neutrophils were already removed above.
CD8+ T cells CD8a+
Regulatory T cells FoxP3+ CD25+
Fig 1H. Intracellular FoxP3 - needs fix/perm, and the FMO must go through the same fix/perm as the full stain.
Eomes+ CD8 T cells Eomes+
Fig 1I. Many canine CD8+ T cells express Eomes.
Proliferating CD8 T cells Ki67+
Fig 1I. A minority are Ki-67+ at rest; this is the proliferation readout.