PhyCLIP (Phylogenetic Clustering by Linear Integer Programming)

Updates

25-Jul-2019: v2.1 - Fixed int64 declaration for inter-platform compatibility; tested on Windows and MAC
16-May-2019: PhyCLIP version 2.0 release
- PhyCLIP is now fully Cythonised. If you have been using PhyCLIP previously, please make sure that Cython is installed (if you are using Anaconda, type conda install -c anaconda cython into your terminal/command prompt)
- ~270x improvement in computational speed for a phylogenetic tree with ~1,200 taxa when compared to version 1.0.
- Decreased peak memory usage by >2x
- --force option available to heuristically force putative subtrees with low divergence (low information) to cluster under the stipulated statistical framework (More information to come in wiki documentation).
- Bug fixes.

Overview

PhyCLIP is an integer linear programming (ILP) approach that optimally delineates a tree into statistically-principled clusters. Other than a rooted phylogeny, 3 additional inputs are required from the user:

Minimum number of sequences (S) that can be quantified as a cluster.
Multiple of deviations (gamma) from the grand median of the mean pairwise sequence patristic distance that defines the within-cluster divergence limit.
False discovery rate (FDR) to infer that the diversity observed for every combinatorial pair of output clusters is significantly distinct from one another.

A manuscript describing PhyCLIP (version 1.0) is published in Molecular Biology and Evolution:
Phylogenetic Clustering by Linear Integer Programming (PhyCLIP) Alvin Xiaochuan Han, Edyth Parker, Frits Scholer, Sebastian Maurer-Stroh, Colin Russell Molecular Biology and Evolution, msz053; doi: https://doi.org/10.1093/molbev/msz053

Full documentation: https://github.com/alvinxhan/PhyCLIP/wiki

We highly encourage that you go through the ENTIRE DOCUMENTATION before starting your analysis.

If you are based in an academic institution, you may go to the Quickstart guide to promptly install and run PhyCLIP using the recommended procedure.

Clustering results of PhyCLIP maximises inclusivity and thus should not be interpreted as taxa linked by rapid transmission events. We reccomend you to use Phydelity, a re-purposing of PhyCLIP for the identification of putative transmission clusters.