Main Content


Construct phylogenetic tree using neighbor-joining method


PhyloTree = seqneighjoin(Distances)
PhyloTree = seqneighjoin(Distances, Method)
PhyloTree = seqneighjoin(Distances, Method, Names)
PhyloTree = seqneighjoin(..., 'Reroot', RerootValue)

Input Arguments

DistancesMatrix or vector containing biological distances between pairs of sequences, such as returned by the seqpdist function.
MethodCharacter vector or string specifying a method to compute the distances between nodes. Choices are 'equivar' (default) or 'firstorder'.

Either of the following:

  • Vector of structures with the fields Header and Name

  • Cell array of character vectors or string vector

The number of elements must equal the number of samples used to generate the pairwise distances in Distances.


PhyloTree = seqneighjoin(Distances) computes PhyloTree, a phylogenetic tree object, from Distances, pairwise distances between the species or products, using the neighbor-joining method.

PhyloTree = seqneighjoin(Distances, Method) specifies Method, a method to compute the distances of the new nodes to all other nodes at every iteration. The general expression to calculate the distances between the new node, n, after joining i and j and all other nodes (k), is given by

D(n,k) =  a*D(i,k) + (1-a)*D(j,k) - a*D(n,i) - (1-a)*D(n,j)

This expression is guaranteed to find the correct tree with additive data (minimum variance reduction).

Choices for Method are:

equivar (default)Assumes equal variance and independence of evolutionary distance estimates (a = 1/2), such as in the original neighbor-joining algorithm by Saitou and Nei, JMBE (1987) or as in Studier and Keppler, JMBE (1988).
firstorderAssumes a first-order model of the variances and covariances of evolutionary distance estimates, with 'a' being adjusted at every iteration to a value between 0 and 1, such as in Gascuel, JMBE (1997).

PhyloTree = seqneighjoin(Distances, Method, Names) passes Names, a list of names (such as species or products), to label the leaf nodes in the phylogenetic tree object.

PhyloTree = seqneighjoin(..., 'Reroot', RerootValue) specifies whether to reroot PhyloTree. Choices are true (default) or false. When RerootValue is false, seqneighjoin excludes rerooting the resulting tree, which is useful for observing the original linkage order followed by the algorithm. By default seqneighjoin reroots the resulting tree using the midpoint method.


collapse all

Create an array of structures representing a multiple alignment of amino acids:

seqs = fastaread('pf00002.fa');

Measure the Jukes-Cantor pairwise distances between sequences.

distances = seqpdist(seqs,'method','jukes-cantor','indels','pair');

Use the output argument distances, a vector containing biological distances between each pair of sequences, as an input argument to seqneighjoin.

Build the phylogenetic tree for the multiple sequence alignment using the neighbor-joining algorithm. Specify the method to compute the distances of the new nodes to all other nodes.

phylotree = seqneighjoin(distances,'equivar',seqs)
    Phylogenetic tree object with 32 leaves (31 branches)

View the phylogenetic tree:



[1] Saitou, N., and Nei, M. (1987). The neighbor-joining method: A new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4(4), 406–425.

[2] Gascuel, O. (1997). BIONJ: An improved version of the NJ algorithm based on a simple model of sequence data. Molecular Biology and Evolution 14 685–695.

[3] Studier, J.A., Keppler, K.J. (1988). A note on the neighbor-joining algorithm of Saitou and Nei. Molecular Biology and Evolution 5(6) 729–731.

Version History

Introduced before R2006a