(It's the most inflammatory of questions!)
What makes dogs breeds so different is that the dog genome is extremely plastic: it has a lot of room for variation as evidenced by how far pugs have come from wolves, even though theoretically they can still mate and produce fertile offspring.
Is there evidence that dogs genome is more plastic than humans?
To my (very limited) knowledge we've not tried selectively breeding humans within bloodlines in anywhere near the same way that breeds of dog have been selected. Are there other domestic animals that have recieved the same degree of breeding but not reached such variation?
Whereas certain breeds of dog are prone to developing certain medical conditions, like hip dysplasia or heart failure, certain ethnicities of human are more likely to develop certain medical conditions, like Tay-Sachs or Sickle Cell Anemia.
To start categorizing humans in such a manner invites similar detachment, with the hazard of conclusions potentially being applied to people.
So no, no real analogy.
I'm thinking about aboriginal communities, Pygmy peoples, or even groups like the Ashkenazis.