Please enable JavaScript.
Coggle requires JavaScript to display documents.
Friendship and Natural Selection., What is expected to see from…
Friendship and Natural Selection.
Fundamental characteristics of friendship.
Evolutionary models.
This suggests there homophily can happen if there is a fitness advantage to same-type interactions.
Candidate gene studies.
Gene variant that shows positive correlation between friends.
Gene variant that shows negative correlation between friends.
Social ties.
Allow for homophily connections to be made.
Genes.
Formation, attributes, and structures.
Reasons for homophily in their genotypes.
Choosing environments that involve people with similar phenotypes influenced by specific genotypes.
Friends are made through geographically proximate or ethnoracially like people who share similar ancestry.
Possibly selected by 3rd parties, but in environments where contact is with similar people.
Actively choosing and keeping of friends with similar genotypes. (homophily)
Actively disregarding or avoiding people with different genotypes.(heterophily)
Homophily & Heterophily advantages.
Homophily:
-Same trait at fixation, causing advantage in every interaction.
More positive correlations between friends over negative.
Contributions to evolutionary fitness across traits.
Positive selection on genes and correlation between friends.
Heterophily: Individuals have same/different traits which allows for dissimilar type interactions to occur.
Experiment for first genome-wide analysis of correlation in genotypes between friends.
GWAS techniques to identify certain theorized patterns across the whole genome. Data from the Framingham Heart Study was used.
Analyzed: 466,608 single-nucleotide polymorphism (SNPs) in 1,932 unique subjects who are in one or more of 1,367 friendship pairs.
Homophily assessment: calculate kinship coefficient (probability two alleles at random from two people are identical)
Heterophily assessment: calculate empirical probability of two individuals having opposite genotypes at a given SNP.
Used for comparison: calculated measures for all nonkin "stranger" pairs from the 1,932 subjects in the friends sample.
Results of genome wide analysis
1,196,429 stranger pairs show kinship coefficients for friends shifted right relative to the strangers.
Friends are more genetically "related" than strangers and the size of difference slightly corresponds to kinship coefficient seen in 4th cousins.
Friends have fewer SNPs where genotypes are opposites.
Pairs of friends are generally more homophilic than pairs of strangers from same population.
GWAS regressing subject's expected genotype on friend's expected genotype.
10 principal components for both homophily and heterophily used as control for ancestry.
907 friend pairs where kinship is ≤0.
Used unimputed and imputed SNPs to improve power.
Average correlation showed slightly negative kinship = -0.003
Variance inflation for friends is more than double, at λ = 1.046,
Shows that polygenic homophily and/or heterophily cause some inflation resulting in large numbers of SNPs being correlated between pairs of friends across genome.
Friends GWAS yields significantly more outliers than the comparison stranger group for both homophily/heterophily.
"Friendship Score"
Sample of 458 friend pairs and 458 stranger pairs.
One-standard-deviation change in the friendship score increases the probability that a pair in the sample are friends by 6% and helps explain 1.4% variance in friendship ties.
Predicts if two people are likely to be friends in a hold-out sample based on genomes.
174 homophilic genes suggest friends ted to have genotypes that yield similar sense of smell. Top 25% homophilic genes have the linoleic acid metabolism system overrepresented.
Heterophilic genes suggests friends have different genotypes for coping with infection.
Top 20% of homophilic SNPs have higher composite of multiple signals (CMS) than the other 80%.
Overall results
Humans tend to share genotypes that are common with their friends and are more likely to be under recent natural selection than other genotypes.
Genotypes positively related under selection show that genes from other people can modify fitness advantages, affecting speed and outcome of evolution.
Reference
Christakis NA, Fowler JH. 2014. Friendship and natural selection. PNAS. 111(Supplement 3):10796–10801. doi:10.1073/pnas.1400825111.
What is expected to see from theoretical models.