Everything is heritable: "Detection of human adaptation during the past 2,000 years", Field et al 2016:

"Detection of recent natural selection is a challenging problem in population genetics, as standard methods generally integrate over long timescales. Here we introduce the Singleton Density Score (SDS), a powerful measure to infer very recent changes in allele frequencies from contemporary genome sequences. When applied to data from the UK10K Project, SDS reflects allele frequency changes in the ancestors of modern Britons during the past 2,000 years. We see strong signals of selection at lactase and HLA, and in favor of blond hair and blue eyes. Turning to signals of polygenic adaptation we find, remarkably, that recent selection for increased height has driven allele frequency shifts across most of the genome. Moreover, we report suggestive new evidence for polygenic shifts affecting many other complex traits. Our results suggest that polygenic adaptation has played a pervasive role in shaping genotypic and phenotypic variation in modern humans."

Graphs: https://twitter.com/KirkegaardEmil/status/729282134272294914

Previously: "Genome-wide patterns of selection in 230 ancient Eurasians" ( http://www.gwern.net/docs/2015-mathieson.pdf ), Mathieson et al 2015. Background on methods: http://www.unz.com/gnxp/the-2000-year-selection-of-the-british/

As I mentioned then, recently on SSC (http://slatestarcodex.com/2016/05/04/myers-race-car-versus-the-general-fitness-factor/), and Cochran emphasizes (https://westhunt.wordpress.com/2016/05/08/you-dont-need-a-weatherman/), as more ancient genomes become available, we will spot more and more signs of selection in the form of soft polygenic sweeps creating between-population differences. Like with finding IQ genes, it is a matter of sample size / power, and as the sample size increases geneticists will be 'surprised' to discover all the selection which has been going on over the past 10000 years.

This paper is particularly neat because it bypasses the slow accumulation of ancient DNA for direct analysis of abundant modern SNP / whole-genome data. As they only use ~3k whole genomes†, we may be able to expect much bigger upcoming finds of selection.

† I am not clear why the sample is so small; they only report signs of selections on SNPs and I'm not sure whether they use whole-genomes simply because they have access to them and not, say, the UK Biobank, or whether the whole-genomes are genuinely needed to do the ancestry/coalescing reconstruction and all the selection is on SNPs because their functions have often been established already in GWASes so they can say 'there was selection on SNP X for trait Y'. EDIT: apparently the method does depend on whole-genomes to find unique mutations according to one of the authors: https://www.reddit.com/r/science/comments/4k1z82/scientists_track_last_2000_years_of_british/d3bzsk0?context=3

"We use 3k genomes (from UK10K) because we need whole-genome data, not genotyping data, to compute our singleton density score (SDS). Singletons are variants private to 1 individual which aren't detectable in genotyping for variants already known; as a result, we are limited to using whole-genome sequencing data. As we get more of that for more populations, we hope to zoom in even closer on evolution - won't that be exciting! :)"