Noooo, they were correlating chunks of the genome to facial features, with a suspiciously large margin of error. They are still way waaaayy distant from being able to alter DNA to give you those facial features.

And as I pointed out elsewhere. There are many many facial features, and shades of distinction amongst those features, and where you draw the lines to declare a race is not objective, "ontologically" or otherwise. The races defined even vary across different disciplines - forensic anthropology versus sociology versus archaeology for example - and are even contested within those disciplines, and are also in a state of flux.