Please enable JavaScript.
Coggle requires JavaScript to display documents.
Friendship and natural selection - Coggle Diagram
Friendship and natural selection
Citation
Christakis NA, Fowler JH. 2014. Friendship and natural selection. Proc Natl Acad Sci U S A. 111(3):10796-10801. doi: 10.1073/pnas.1400825111
Introduction
Background
Similarities shown in members of friendships
Possible reasons for homophily
Similar locations/ethnicities create passive shared ancestry
Similarities attract "birds of a feather flock together"
Similar genotypes actively choosing similar environments
External factors that gather similar people
Possible reasons for heterophily
Environments where different traits are desirable may group dissimilar people
"Opposites attract"
Hypothesis
Genomes of friends will be more similar than dissimilar
Friends display greater genetic similarities over genome as a whole
Methods
genome-wide association study (GWAS)
Not testing for friendliness
Analyzing patterns across genome
1,932 individuals & 1,367 friendships
Not testing specific single-nucleotide polymorphisms (SNPs)
1,468,013 common SNPs
split-sample replication study
False positive failsafe
Used 458 pairs of friends that fell outside of the kinship range for GWAS (higher than zero)
Kinship coefficients calculated
Probability of randomly sampled alleles between two people matching
Genome-wide complex trait analysis (GCTA) method
Positive values = positive correlation of genotypic similarity
Negative values = opposite genotype
For comparison
1,196,429 stranger pairings
Ran all tests on "strangers" among all applicants
Friends and strangers GWAS contained same people with different relationships
Difference-in-means test
Ensured friends were unrelated
No control against population stratification
Ethnically similar
Results
Difference-in-means test
Friends were genetically more similar than strangers of same population
Friends have higher kinship coefficients
Friends had lower proportion of opposite genotypes
Split-sample replication study
"Friendship score" increased probability of compared samples to be friends by 6%
GWAS
Control against population stratification
No subjects flagged as outliers
Controls for ancestry left in
Control against distant relatives: use only kinship less than or equal to zero.
Friends showed more homophily and heterophily than strangers
Polygenic homophily and/or heterophily more likely than chance
Statistically significant correlations of SNPs
Gene-set anaylsis
Top 1%: Similar sense of smell
Top 25%: Similar linoleic acid metabolism
Top 25%: Difference in immune system
No significant overrepresentation in strangers
Discussion
Implications
"Friendship score" based on genome relatedness
Friendships and romantic relationships could be influenced by genetic structure
Genes cause people to seek out similar social & physical
environments?
Limitations
Distant relatives control creates bias against homophily
Ethnic similarities
Further Quations
How does this relate to the social interactions of people who are closer to their friends than their family
Why is olfactory similarities so linked
How much do genes control our search for certain environments