TY - JOUR
T1 - PhyloMeasures
T2 - A package for computing phylogenetic biodiversity measures and their statistical moments
AU - Tsirogiannis, Constantinos
AU - Sandel, Brody Steven
PY - 2016/7/1
Y1 - 2016/7/1
N2 - We present PhyloMeasures, a new software package including both a C++ and R version, that provides very fast computation of popular phylogenetic diversity measures. PhyloMeasures introduces two major advances over existing methods. First, it uses efficient algorithms for calculating basic phylogenetic metrics (such as Faith's PD and the mean pairwise distance, MPD) and two-sample measures (such as common branch length, CBL, and the unique fraction) that are designed to perform well even on very large trees. Second, it computes exact richness-standardised versions of these measures (such as the widely used net relatedness index, NRI) by efficiently evaluating analytical expressions for the mean and variance of the basic measures, rather than by the slow and inexact randomization techniques that are the current standard. Together, these lead to massive improvements in performance compared to the current state of the art. For example, running on a standard laptop, PhyloMeasures functions can provide the NRI for 20 samples from a tree of 100 000 tips in about 1.5 s, compared to an estimated 37 d using standard resampling approaches. This will allow analyses on larger data sets than were previously possible.
AB - We present PhyloMeasures, a new software package including both a C++ and R version, that provides very fast computation of popular phylogenetic diversity measures. PhyloMeasures introduces two major advances over existing methods. First, it uses efficient algorithms for calculating basic phylogenetic metrics (such as Faith's PD and the mean pairwise distance, MPD) and two-sample measures (such as common branch length, CBL, and the unique fraction) that are designed to perform well even on very large trees. Second, it computes exact richness-standardised versions of these measures (such as the widely used net relatedness index, NRI) by efficiently evaluating analytical expressions for the mean and variance of the basic measures, rather than by the slow and inexact randomization techniques that are the current standard. Together, these lead to massive improvements in performance compared to the current state of the art. For example, running on a standard laptop, PhyloMeasures functions can provide the NRI for 20 samples from a tree of 100 000 tips in about 1.5 s, compared to an estimated 37 d using standard resampling approaches. This will allow analyses on larger data sets than were previously possible.
UR - http://www.scopus.com/inward/record.url?scp=84945157533&partnerID=8YFLogxK
U2 - 10.1111/ecog.01814
DO - 10.1111/ecog.01814
M3 - Journal article
AN - SCOPUS:84945157533
SN - 0906-7590
VL - 39
SP - 709
EP - 714
JO - Ecography
JF - Ecography
IS - 7
ER -