6/ This work was led by Keith D. Harris, with Yuval Talmor and Meirav Yefe Nof making important contributions, and excelent input provided by Nimrod Marom, Yitzchak Jaffe and Viviane Slon.

5/ Results will (shortly) be available for visualization and analysis via the interactive web interface DORA (dora.modelrxiv.org) for further exploration (we'll post here when its ready).

4/ To enable the analysis, we developed cool new methods: (1) impHet, a pipeline that provides high accuracy estimates of heterozygosity for ancient genomes, (2) a new genomic scan for balancing selection in aDNA timeseries data, (3) a simulation pipeline for mimicking biases in aDNA.

3/ We also characterize "peaks" in disease burden, with a major peak at 7000BP in both regions, and a smaller, later peak corresponding to urbanization and the rise of large polities. Interestingly, increases in diversity of immune-related genes is highly correlated between the regions (r=0.91)

2/ We found clear evidence that the answer is YES. We show that the MHC is the genomic region with the most dramatic increases in heterozygosity - 16 and 7 times higher than genome-wide levels, in two regions, Southwestern Eurasia and East Asia). 66% of Het-increasing genes are in the MHC.

Perhaps the main moral from this story - catagorical social classes are certainly useful, but as we accumulate more types of data on more specie, it is worthwhile to let the data do the talking, and quantify composite sociality phenotypes directly from mulitdimensioal data.

We also looked at genomic signitures in 12 candidate genes in relation to our composite social phenotype, and we see that the transition cannot be explained by directional selection on a specific gene. More likely, the transition to high sociaility in bees involved a combination of genetic changes.

The period right after the transition was the one with the largest movments in the phenotypic space. Importantly, although "eusocial" is often treated as a single socialility class, the high social complexity region in our analysis was larger than that of all other lower sociality phenotypes.

The transition path seems different than that of the "social ladder" theory, where extant social complexity catagories represent steps towards high social complexity.

Looking at the movment in this phenotypic space, we see a major transition around 70mya, which breached to a new region in the space. This transition was followed by a period of divesification that resulted in a wide range of high-complexity phenotypes of honey bees, stingless bees, and bumble bees.

We decided not to take traditional social catagories as an a priori assumption, and instead generated a quantitative social complexity phenotype from 17 social-related traits for 80 bee species. We then tracked the evolutionary history in PCA space of the composite social complexity phenotype.

I would like to be added. Thanks.