odmix: Gaussian mixture modelling of distances in a multiple...

In odseq: Outlier detection in multiple sequence alignments

Description

This function performs clustering of biological sequences via fitting a Gaussian mixture model of the distances defined by the odseq algorithm

Usage

odmix(msa_object, distance_metric, groups)

Arguments

msa_object	An object of formal class MsaAAMultipleAlignment, as provided by the msa package.
distance_metric	A string indicating the type of distance metric to be computed. Either 'linear' and 'affine' is supported at the moment.
groups	Number of groups to fit in the mixture model. If a numeric vector of size n, n models will be fitted and a list of BIC values will be given to choose a single model.

Value

A list containing the following items:

prob	A numeric matrix of size n x groups where the probability of belonging to a group is provided for each sequence.
class	The class assigned according to prob. Returns a numeric vector.
BIC	BIC values for the models proposed in groups

Author(s)

Jos<c3><a9> Jim<c3><a9>nez <[e-mail]>

See Also

odseq_unaligned odseq

Examples

library(msa)
data(seqs)
al <- msa(seqs)
odmix(al, distance_metric = "affine", groups = 2)

odseq documentation built on Nov. 8, 2020, 6:50 p.m.