Best Pet DNA TestsOne dog, two DNA tests: what Embark and ProsperK9 said about the same Labradoodle
Cosmo, a male Labradoodle owned by this site's founder, was tested by Embark Breed + Health on 2020-12-02 and swabbed for ProsperK9's whole-genome test the same year, at age 3. The ProsperK9 report has been re-analysed since from that stored genome, and the version compared here is the public demo as read on 2026-10-12. Both reports are public: the Embark PDF and the ProsperK9 demo report. Every figure below is copied from one of them.
Owned by the founder of ProsperK9, one of the tests compared here. No affiliate links. Facts read on each company's product page on 2026-10-09. Jump to the table · About the site
Read this first. One dog is a demonstration, not a study: it shows what each report contains and where the two agree, and it cannot say which test is more accurate. The dog belongs to the owner of this site, who also founded ProsperK9, so the ProsperK9 column is the one to check against the demo yourself.
Breed mix
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| Poodle (all sizes) | 75.1% (Standard 40.2%, Small 34.9%) | 75.7% (Standard 38%, Miniature 21%, Toy 17%) |
| Labrador Retriever | 10.4% | 10.1% |
| Cocker Spaniel | 9.4% | 6.5% |
| English Cocker Spaniel | 5.1% | 4.7% |
| Barbet | not reportedThe one disagreement. Barbet is a French water dog closely related to the Poodle; Embark's 2020 panel folds that ancestry into Poodle. | 3.0% |
| Chromosome painting | Yes, all 38 pairs | Yes, all 38 pairs; 58% of the genome assigned with confidence, the rest shown lighter where two related breeds fit equally |
Two technologies, swabs from the same year, agreement to within 3 percentage points on every breed. Embark splits Poodle into Standard and Small; ProsperK9 into Standard, Miniature and Toy.
Ancestry lines, inbreeding and immune diversity
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| Maternal haplogroup | B1 (haplotype B84) | B; closest named haplotype B1, 11 mutations away; sequenced at 234x |
| Paternal haplogroup | A1a (haplotype H1a.59) | A; sub-branch (A1a etc.) not resolved by this analysis |
| Inbreeding | 10%Different measures: Embark reports a coefficient from runs of homozygosity on its chip; ProsperK9 reports a genotype-based F against a 1,929-dog panel. They are not directly comparable. | 17.0% (33.8th percentile of 1,929 dogs) |
| Immune-gene (MHC) diversity | High diversity at DRB1 and at DQA1/DQB1 | High: class II at the 80th percentile, class I at the 63rd |
Genetic health
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| Known variants tested | 197 (25 breed-relevant, 172 other) | 183, plus MDR1, merle and a mitochondrial screen |
| Two copies (at risk) | 0 | 0 |
| One copy (carrier) | 0 reportedCopper toxicosis was not on Embark's 2020 panel, so the ATP7B result cannot be compared. BMP3 appears in Embark's traits section as AC, the same heterozygous result. | 2: ATP7B (copper toxicosis) and BMP3 (no known effect) |
| MDR1 drug sensitivity | Clear | Normal, no copies |
| Whole chromosomes checked | no | All 38 pairs plus X present in the expected number; two 1-Mb duplications at chromosome ends that 3 to 4% of reference dogs share |
| Large copy-number changes | no | 2 rare: a 0.15-Mb deletion on chr 21 across olfactory-receptor genes, a 0.20-Mb duplication on chr 24; neither linked to disease |
| Rare predicted gene-disrupting variants | no | 4 flagged (TECPR2, LRP4, CHD4, CACNA1G), each carried by 1 to 4% of dogs, labelled as computational predictions, not tested conditions |
| Hip and cruciate risk from ancestry | no | Hip dysplasia 11.5% (population about 12%); cruciate tear 0.99x a typical mixed breed |
Both tests found Cosmo clear of every two-copy condition they looked for. Beyond the known-variant list, the rows marked 'no' for Embark are kinds of analysis a marker chip does not do.
Coat and body genes
Of the 21 single-gene results both tests report with a definite call, 20 agree. The exception is shedding (MC5R). ProsperK9 leaves 5 loci unresolved or low confidence at its sequencing depth rather than guessing; Embark's chip reads those directly.
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| E locus (MC1R) | ee | e? e?, low confidence |
| B locus (TYRP1) | Bb | B/b, medium confidence |
| D locus (MLPH) | DD | D/D |
| K locus (CBD103) | KBKB | KB/KB |
| A locus (ASIP) | atat | not resolved |
| S locus (MITF) | SS | not resolved |
| Merle (PMEL) | mm | m/m, read directly from 2 reads |
| Furnishings (RSPO2) | FF | FF |
| Coat length (FGF5) | LhLh | LhLh |
| Coat texture (KRT71) | CC | CC |
| Shedding (MC5R) | CTThe one disagreement among comparable single-gene results. | CC |
| Muzzle length (BMP3) | AC | AC |
| Tail length (T) | CC | CC |
| Hind dewclaws (LMBR1) | CC | CC |
| Blue eyes (ALX4) | NN | NN |
| Body size: IGF1 | NN | NN |
| Body size: IGF1R | GA | AG |
| Body size: STC2 | TT | not reported |
| Body size: GHR E191K | GA | AG |
| Body size: GHR P177L | CC | CC |
| Altitude (EPAS1) | GG | GG |
| Appetite (POMC) | NN | not resolvable at this depth |
| Hairlessness (FOXI3) | NN | NN |
| Albinism (SLC45A2) | NN | NN |
| Saddle tan (RALY) | II | II |
| Red pigment intensity | Intermediate | Intermediate (3 of 8 intensifying alleles) |
Size and other predictions
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| Predicted adult weight | 33 lbCosmo actually weighs 29 lb (owner's record, 2026-10-12): Embark 4 lb over, ProsperK9 4.3 lb under. Different models; ProsperK9's combines breed ancestry with a genome-wide size score. | 24.7 lb (11.2 kg) |
| Actual adult weight | 29 lbThe one row a reader can check against reality. Both predictions miss by about the same amount in opposite directions. | 29 lb |
| Predicted height | not reported | 41.4 cm at the withers |
| Wolfiness | 2.8%, labelled high | not reported as such; wolf is in the ancestry panel and was not detected in Cosmo |
Oral microbiome
| Embark tested 2020-12-02 | ProsperK9 swabbed 2020, demo read 2026-10-12 | |
|---|---|---|
| Oral microbiome | not offered | Read from the same swab |
| Microbiome age | 4.2 years against 3.0 years recorded at the swab, i.e. the mouth bacteria resemble an older dog's | |
| Gum-disease bacteria | 26.9% of bacterial reads, 59.7th percentile; five Porphyromonas and Tannerella species named with their shares | |
| Bacterial activity | Breath sulfur 30th percentile, gum-damaging enzymes 52nd, nitrate recycling 57th, ammonia 58th; summarised as typical |
Only ProsperK9 reads the mouth bacteria. These results describe the swab on the day it was taken and change between swabs.
Only in the Embark report
- A family tree (parents, grandparents, great-grandparents) inferred from the breed mix.
- A dog-relatives finder and a breed-reveal video.
- The 2020 panel tested 14 more known variants than ProsperK9's current catalogue (197 against 183).
Only in the ProsperK9 report
- Everything in the oral microbiome section.
- Chromosome-level and large copy-number checks, and the rare predicted gene-disrupting variants.
- Ancestry-weighted hip and cruciate-ligament risk.
- Per-result confidence grades and sequencing depth, and the honest 'not resolved' where depth was insufficient.
- A Healthbook holding the report alongside records, with an AI Health Advisor and a vet summary.
What this does and does not show
- It shows the two technologies converge on the same dog. Breed mix within 3 points, the same maternal and paternal families, the same high immune diversity, clear on every two-copy condition, and 20 of 21 comparable coat and body genes identical.
- It shows what whole-genome sequencing adds and what it costs. Added: chromosome and copy-number checks, rare variants, the microbiome, and confidence on every call. Cost: a handful of loci left unresolved at this sequencing depth that a chip reads outright, and a shedding call that differs.
- It does not show accuracy. Agreement between two tests is not correctness, and one dog is not a sample. Neither report was checked against Cosmo's pedigree, because a Labradoodle has none.
- It does not compare today's Embark. Both swabs are from 2020. Embark's report is the one delivered then; its panel and breed list have grown since. ProsperK9's is the current re-analysis of the 2020 genome, which is the point of keeping a genome on file, but it also means the two reports are not the same age.
Sources
- Embark Breed + Health report for Cosmo, test date 2020-12-02, re-downloaded 2026-04-30: PDF, 37 pages
- ProsperK9 public demo report (Cosmo, swabbed 2020 at age 3), read 2026-10-12: https://my.prosperk9.com/demo
Embark vs ProsperK9, the product comparison · All four tests · How we compare