Erin Fry's selection for Journal Club: The phenotypic legacy of admixture between modern humans and Neandertals

Erin Fry

Second year Graduate Student

I presented "The phenotypic legacy of admixture between modern humans and Neandertals" (Simonti et al., Science, 2016) on 5/4/2016.

Anatomically modern humans (AMHs) left Africa about 60,000 years ago. In their migrations to Europe and Asia, AMHs encountered and cohabited with at least two ancient hominin species - the Neanderthals and the Denisovans - until as recently as 30,000 years ago.  During this period of cohabitation, AMHs admixed, or exchanged genetic material through interbreeding, with these ancient hominid species, leaving specific genetic signatures in the genomes of AMHs. Recent work estimates that 1.5-4% of modern Eurasian genomes are a product of admixture between AMHs and Neanderthals (Reviewed by Vattathil and Akey, 2015).


Two prominent papers published in 2014 (Sankararaman et al. and Vernot and Akey) reported signatures of positive and negative selection on Neanderthal-derived alleles throughout the genome. Introgression of Neanderthal alleles into European and Asian genomes varies dramatically across the genome, creating gene ‘deserts’ which completely lack Neanderthal-derived alleles. While some regions have very little evidence of admixture, some Neanderthal alleles appear to have been advantageous for ancestral AMHs. For example, Neanderthal alleles at high frequency in modern humans are enriched for genes implicated in skin phenotypes. These findings led to the hypothesis that Neanderthals contributed alleles to ancestral AMH populations that were adaptive for living at higher latitudes and in non-African climates. These alleles were hypothesized to contribute to complex traits like skin pigmentation, lipid metabolism, mood disorders, and pathogenic resistance. As our department discussed the report by Simonti et al. we were keen to examine the evidence for or against this hypothesis as well as evaluate the results from the first study to utilize electronic health data in conjunction with introgression-informed genetic analyses.


Simonti et al. utilize electronic health record (EHRs) data from eMERGE to identify medical phenotypes affected by introgressed Neanderthal alleles in modern individuals of European descent. They first integrated genotypic data from eMERGE with the Neanderthal genome sequence to identify Neanderthal-derived alleles in each individual. They then employed a variety of genetic association approaches to understand the contribution of these alleles to various phenotypes determined by the EHR ICD-9 billing codes. These EHRs have tremendous potential, but due to complexities in medical diagnoses, EHRs may be uninformative at best and misleading at worst. Our department was concerned that the extensive quality control was impossible to achieve for a study of this size (28,146 individuals). As networks like eMERGE continue to grow and improve, we anticipate that despite classification difficulties, these data will be extremely valuable.


Utilizing GCTA, they tested the overall genetic contribution of Neanderthal alleles to 46 medical traits. After including a non-Neanderthal-derived genetic relationship matrix in their model and a replication phase, Simonti et al. found three phenotypes significantly influenced by Neanderthal alleles: actinic keratosis, mood disorders, and depression (Table 1). They reported that Neanderthal-derived alleles contributed to individual risk of developing these conditions, however, as seen in a supplemental table, only half of Neanderthal-derived SNPs are risk inducing. Simonti et al. concluded that Neanderthal-derived alleles contribute to these three phenotypes. However, our department considered the data under the adaptive allele hypothesis, as well. We concluded that these results do not provide strong support for the hypothesis, possibly because the methods employed in the paper are not appropriate or powerful enough to specifically address that question.


Their next genetic association method, PheWAS, associates traits defined by ICD-9 disease codes, symptoms, causes of injury, and medical procedures, with specific genetic variants. After examining 1,152 phenotypes for association with Neanderthal alleles, Simonti et al. identified four traits associated with one Neanderthal-derived SNP each. They evaluate each association, considering nearby genes implicated in the phenotype as well as histone modification and eQTL data. Their most compelling argument for functional plausibility is for Hypercoagulable state’s association with rs3917862. The SNP is an eQTL for two nearby genes implicated in leukocyte recruitment and blood coagulation and bears regulatory histone modifications in blood and epithelial cells. However, without functional validation of this SNP, or the other three associated SNPs, there is no solid evidence for their effect on downstream gene expression or phenotypic consequences.


The evidence presented in this paper suggests that Neanderthal-derived alleles have a small effect on modern human phenotypes, including depression and skin pigmentation. Our department concluded that their findings do not provide evidence for or against the adaptive allele hypothesis, but future work that utilizes similar EHRs may. More broadly, this report was the first to utilize EHRs and signatures of admixture between AMHs and Neanderthals to provide evidence for phenotypic consequences of Neanderthal-derived alleles modern Europeans. 

Editor: Anna Di Rienzo, Professor, Dept of Human Genetics