iNJclust: Iterative Neighbor-Joining Tree Clustering Framework for Inferring Population Structure

Abstract

Understanding genetic differences among populations is one of the most important issues in population genetics. Genetic variations, e.g., single nucleotide polymorphisms, are used to characterize commonality and difference of individuals from various populations. This paper presents an efficient graph-based clustering framework which operates iteratively on the Neighbor-Joining (NJ) tree called the iNJclust algorithm. The framework uses well-known genetic measurements, namely the allele-sharing distance, the neighbor-joining tree, and the fixation index. The behavior of the fixation index is utilized in the algorithm's stopping criterion. The algorithm provides an estimated number of populations, individual assignments, and relationships between populations as outputs. The clustering result is reported in the form of a binary tree, whose terminal nodes represent the final inferred populations and the tree structure preserves the genetic relationships among them. The clustering performance and the robustness of the proposed algorithm are tested extensively using simulated and real data sets from bovine, sheep, and human populations. The result indicates that the number of populations within each data set is reasonably estimated, the individual assignment is robust, and the structure of the inferred population tree corresponds to the intrinsic relationships among populations within the data.

Nonparametric approaches for population structure analysis

“…Furthermore, NETVIEW [ 63 ] reveals the hierarchy of population substructures based on a representation of the genetic data as a network of individuals connected by edges representing the ASD between each pair. Iterative neighbor-joining tree clustering (iNJclust) [ 64 ] performs a graph-based clustering on a neighbor-joining (NJ) tree. Table 2 describes the distance-based methods in terms of the proximity measure, clustering technique, and available package/tool.…”

“…(2012) [ 63 ] Super paramagnetic clustering (SPC) NETVIEW [ 86 ]: MATLAB Limpiti at el. (2014) [ 64 ] Neighbor-joining (NJ) tree-based clustering iNJclust [ 87 ]: C++ …”

A different view on fine-scale population structure in Western African populations

“…STRU CTU RE and ADMIXTURE are used to determine how individuals are inherited from a certain number of population ancestries (K) using maximum likelihood estimation from SNPs (Alexander et al 2009), while PCA refers to a relatively small number of uncorrelated variables derived from an initial pool of variables while explaining as much of the total variance as possible. A higher resolution in population clustering is only obtained when fine-scale structure detection tools are applied, including haplotype-based clustering (e.g., fin-eSTRU CTU RE jointly with CHROMOPAINTER (Lawson et al 2012) and iterative pruning method for clustering [e.g., iNJclust (Limpiti et al 2014), and SHIPS (Bouaziz et al 2012), and ipPCA (Intarapanich et al 2009)]. fineSTRU CTU RE and CHROMOPAINTER have been already applied to the African context (Busby et al 2016;Patin et al 2017) and solved Western African clustering to a fine-scale magnitude, showing that most sub-Saharan populations share a certain proportion of ancestry with groups from outside of their current geographic region (sharing between different ethnolinguistic groups, for example, western Bantu speakers having some input from western Pygmies) as a result of gene-flow within the last 4000 years.…”

ipADMIXTURE: R package for inferring sub-population clusters based on genetic admixture

“…There are various methods that can perform data clustering on genetic data and report phylogeny of clusters, such as ipPCA [6], iNJclust [7], IP-CAPS [8], etc. However, no methods use admixture analysis that contains information of common ancestry to perform clustering.…”