Genetic diversity in global chicken breeds as a function of genetic distance to the wild populations

Migration of populations from their founder population is expected to cause a reduction in genetic diversity and facilitates population differentiation between the populations and their founder population as predicted by the theory of genetic isolation by distance. Consistent with that, a model of expansion from a single founder predicts that patterns of genetic diversity in populations can be well explained by their geographic expansion from the founders, which is correlated to the genetic differentiation. To investigate this in the chicken, we have estimated the relationship between the genetic diversity in 172 domesticated chicken populations and their genetic distances to wild populations. We have found a strong inverse relationship whereby 87.5% of the variation in the overall genetic diversity of domesticated chicken can be explained by the genetic distance to the wild populations. We also investigated if different types of SNPs and genes present similar patterns of genetic diversity as the overall genome. Among different SNP classes, the non-synonymous ones were the most deviating from the overall genome. However, the genetic distances to wild populations still explained more variation in domesticated chicken diversity in all SNP classes ranging from 81.7 to 88.7%. The genetic diversity seemed to change at a faster rate within the chicken in genes that are associated with transmembrane transport, protein transport and protein metabolic processes, and lipid metabolic processes. In general, such genes are flexible to be manipulated according to the population needs. On the other hand, genes which the genetic diversity hardly changes despite the genetic distance to the wild populations are associated with major functions e.g. brain development. Therefore, changes in the genes may be detrimental to the chickens. These results contribute to the knowledge of different evolutionary patterns of different functional genomic regions in the chicken. Author summary The chicken was first domesticated about 6000 B.C. in Asia from the jungle fowl. Following domestication, chickens were taken to different parts of the world mainly by humans. Evolutionary forces such as selection and genetic drift have shaped diversification within the chicken species. In addition, new breeds or strains have been developed from crossbreeding programs facilitated by man. These events, together with other breeding practices, have led to genomic alterations causing genetic differentiation between the domesticated chickens and their ancestral/wild population as well as manipulation of the genetic diversity within the domesticated chickens. We investigated the relationship between 172 domesticated chicken populations from different selection, breeding and management backgrounds and their genetic distance to the wild type chickens. We found that the genetic diversity within the populations decreases with the increasing genetic distances to the wild types. Human manipulation of chicken genetic diversity has more effect on the genetic differentiation than simple geographic separations (through migrations) do. We further found that some genes associated with vital functions show evolutionary constraints or persistent selection across the populations and do not comply with this relationship i.e. the genetic diversity within the populations is constant despite the change in the genetic distance to the wild types.

the 'Out of Africa' theory which asserts that modern humans originate from Africa [13] and 93 human populations worldwide resulted in a reduction in genetic diversity with the increasing 94 geographic distance from east Africa (Ethiopia) [4,5,14,15]. Similar studies in cattle also 95 reported a decreasing genetic diversity with increasing geographic distance to the cattle 96 domestication center in Southwest Asia [16,17]. 97 The loss of genetic diversity within the migrated populations, which can be explained by the 98 geographic distance from their founders, is believed to be a good measure of neutral genetic 99 diversity as a consequence of genetic drift. However, the overall genetic diversity is also a result 100 of population specific events such as mutations, natural selection to favor adaptation in the 101 current environments and/or artificial selection (e.g. in livestock production practices) as well as 102 population specific drift [5]. Consequences of selection are often measured by non-neutral 103 genetic variation as it is assumed that non-neutral regions with functional fitness effects in the 104 genome evolve differently to the neutral genome. In this study we used the global collection of 105 chicken breeds [7] to investigate the pattern of the overall genetic diversity moving outwards the 106 centers of chicken domestication, given all events taking place in the genome. Furthermore, we 107 investigate if different functional regions of the genome present similar patterns as the overall 108 genome. We hypothesized that changes in genetic diversity may be faster in some genes or 109 functional categories depending on their functions and changes may also be different in different 110 breeds or breed groups due to different adaptive or artificial selection targets. Therefore, the 111 pattern of relationship between genetic diversity and genetic distance may behave differently, 112 less complying with the overall genome and more dynamic than the non-genic regions due to 113 differences in selection patterns in addition to other population specific events. Studying the theory of genetic isolation by distance and/or the concept of migration from a single 115 location with chickens poses some challenges because the physical locations do not always 116 represent their geographic origin (following migration from founders). For many chicken breeds 117 the time point when they have migrated to their current locations is unknown. We also believe 118 that geographic distances may not be the best predictor of the genetic diversity in the chicken.

119
This is because unlike in humans where genetic evolution is mostly driven by natural 120 circumstances, rapid migration, crossbreeding forced by man, refined breeding programs and 121 artificial selection for desired traits have largely shaped the evolution of domesticated chickens.

122
The changes in genetic diversity and evolutionary rates are often rapid in domesticated livestock 123 and the genetic architecture of chickens around the same geographic location may also differ 124 greatly depending on different breeding practices or selection targets. Therefore, in our study we  and S1 Table are represented by symbols of different colours and shapes. There is a strong 137 inverse relationship between the genetic diversity within populations and their genetic distances 138 to the wild populations. This relationship is similar even when using just neutral markers 139 (intergenic SNPs, Fig 2). Across these chicken populations, 87.5% (Table 1)  SNPs was almost non-existing with an R 2 value of 0.01. We also used the fixation index ( ) as  i. breeds of the same geographic origin are found scattered across the genetic diversity 183 spectrum. This is the case for Asian (red symbols) and European (green symbols) type 184 breeds. As it is shown in Fig 1 and  ii. the concept of isolation by distance assumes that individuals from nearby locations are likely 196 to be related due to mating possibilities. This is often the case in traditional breeding systems 197 but it is not the case with the fancy and commercial breeding and management practices.  Table   241 1. UTR5 and UTR3 refer to the 5' and 3' UTR classes, respectively.    Non-synonymous. The mean heterozygosities of the SNP classes were significantly different to 295 the overall mean (Welch two sample t-test p < 0.05) except for the 3' UTR and 5' UTR classes.

296
The standard errors (SEs) of the means were lower than 0.005 in all the SNP classes and the 297 overall except for the 5' UTR with SE = 0.009. Different letters in the bars means that there is 298 significant difference in the mean heterozygosity within the same level, e.g. difference between 299 'Non-genic' and 'Genic' classes on the first level or difference between 'Non-synonymous' and 300 'Synonymous' classes on the third level. PTPRS and RTN4) in the lowest 5% were associated with brain development.

319
Genes in the lowest 5% had slopes ranging from -0.110 to -0.319 while the top 5% ranged from -320 0.960 to -1.099 (S3 Table). The genes in the top 5% indicate rapid changes in genetic diversity 321 due to the genetic distance of the chicken breeds to G. gallus while those in the lowest 5% 322 indicate genetic diversity changes at a very slow rate in relation to the genetic distance. We 323 obtained the individual gene functions for these genes in the lowest and top ranges from DAVID 324 annotation platform (S3 Table). The figures showing the relationship between genetic diversity 325 and genetic distance in these genes are shown in S1 File and S2 File for the top and lowest 5% 326 ranges, respectively.

327
The genes in the top 5% slope range were associated with transmembrane transport (SLC25A6, and functional processes, is assumed to be highly conserved in chicken as well as in humans and 352 was reported to be under very strong evolutionary constraint [28]. Other than some of the genes, 353 which are mentioned above for being related to the development of the brain, genes in the lowest 354 5% range were also found to be associated with other important developmental processes, which were based on their continent of origin and/or type as described in S1  coefficients of the overall pattern to the patterns of the different SNP classes.

456
For the individual genes, because some of the genes were annotated with only one or very few 457 associated SNPs while others were annotated with more, we only considered genes with at least 458 ten associated SNPs (resulting in 6 303 in total) for making comparisons with the overall pattern. 459 We evaluated the rate of change in the genetic diversity within the genes due to the change in