Thread by @andersbrgstrm

Our wolf paper is out! nature.com/articles/s4158… Analysing 72 ancient 🐺 genomes from the last 100,000 years, we: #1 chart wolf natural history through the Ice Age, #2 directly detect natural selection, #3 reveal that dogs have dual ancestry. A 🧵 (11), illustrations by @jessrpeto

Wolf history: Throughout the Ice Age, wolf genomes cluster by time, not by geography—implying continuous homogenization of ancestry at a global scale. Genetic differentiation was remarkably low (FST < 3%)—an order of magnitude lower than today—demonstrating global connectivity.

Despite the pervasive gene flow, small amounts of deep, local ancestry persists in Europe and other Eurasian regions, meaning that wolf populations here did not go locally extinct. The wolf species thus seem to have fared well during the dramatic climate changes of the Ice Age.

Selection: The deep genomic time series enabled a true ancient DNA scan for natural selection, across ~30000 generations. Testing every variant for a correlation to time (literally a #gwas for sample age), we identified genomic regions where allele frequencies shift dramatically.

In a few cases, alleles go from ~0% to 100% frequency, and are fixed in all wolves and dogs today. This is thus not just local, but global, species-wide adaptation—to our knowledge the first time this is detected in a macro-organism ‘directly’ from ancient DNA.

The most dramatic selection occurs in the gene IFT88, in the period 40-30 kya. This gene is involved in craniofacial and mandibular development. We speculate that this selective sweep in wolves drove adaptation to new prey or hunting strategies in the changing Ice Age ecosystem.

Dog origins: The origin of dogs remains one of the most elusive questions of natural and human history. We think the way to tackle it is to place dogs into the context provided by wolf ancestry history. Our wolf dataset now allows us to make solid progress on this question.

We find that, overall, dogs are genetically closer to ancient wolves from eastern Eurasia, than to those from western Eurasia. Domestication thus likely happened somewhere in the east—although outside of northeastern Siberia, where sampled ancient wolves don’t match dog ancestry.

However… we also find that a single wolf progenitor cannot explain dog ancestry globally. Instead, Near Eastern and African dogs have some extra attraction to ancient wolves from western Eurasia—even to wolves that lived >28kya, before dogs were likely domesticated.

The best available match for the second source of dog ancestry is not ancient wolves from Europe, but rather present-day ones from the Near East (no ancient genomes from here yet). This second source contributed up to half of the ancestry of early Near Eastern and African dogs.

Our results imply either 1) independent domestication, and merging in the west, or 2) domestication once in the east, followed by gene flow from wild wolves. We cannot distinguish these scenarios currently. Either way, dogs today derive ancestry from at least two wolf sources.

Thanks! This was an immensely collaborative project involving colleagues from 18 countries, 45+ archaeologists and 9 #aDNA labs. Huge thanks to @Nibbledtodeath @UlrikeTaron @MhsSinding @ErikErsmark @LrFrantz @ArchaeOphelie @daniMfernandes @MorganeOllivier @leo_speidel …

@TatianaFireborn @Pooja_Swali @AnnaLinderholm1 @RuthFCarden @DidDigDogs @JenBiologist @AndersGother @PinhasiL @bonesandbugs @Greger_Larson @love_dalen @pontus_skoglund @uni_tue @unipotsdam @LFGeoGenetics @UCSC_PGL @PalaeoBARN @MPI_EVA_Leipzig @CpgSthlm @TheCrick and many others!