PartitionFinder2 on XSEDE

PartitionFinder2 is a program for selecting best-fit partitioning schemes and models of evolution for nucleotide, amino acid, and morphology alignments. The user provides an alignment, and optionally some pre-defined data blocks (e.g. 9 data blocks defining the 1st, 2nd and 3rd codon positions of 3 protein-coding genes, see Figure 1). The program then finds the best partitioning scheme for this dataset, at the same time as selecting best-fit models for each subset of sites/columns.

Manual for PartitionFinder2 [PDF]

PartitionFinder Google Group

Input files: PartitionFinder2 accepts as input two files: one file contains the data set in Phylip format. The second is a configuration (.cfg) file that sets the parameters for the job run. The configuration file can be (or at least would generally be) created by a local version of PartitionFinder2, and uploaded. In theory you can create it manually as well. If you choose either option, please be sure the .cfg file specifies alignment=infile.phy . The .cfg file can also be created using the CIPRES interface. You have to choose between the two options, if you provide your own .cfg file, most interface options will be disabled.

Output files: PartitionFinder2 returns a set of files contained in a directory. The output directory will be zipped up into a single archive file for you to download, unzip, and examine.

best_schemes.txt has information on the best partitioning scheme found and the settings used to find it. This includes a detailed description of the scheme as well as the model of molecular evolution that was selected for each subset in the scheme. It also contains a descriptionof the each scheme in RAxMLand Nexus formats.

subsets folder contains the results of the model selection on each subset of sites that was analysed. These are .txt files, in which each model you included in your analysis is listed, in order of increasing AICc score (i.e. best model is at the top). The default is to save model selection results of only those subsets that made it into the best partitioning scheme. If you want to save the model selection results from all subsets of sites that were analysed, then you can use the save phylofiles commandline option (but beware, this results in writing a lot of files see info on the option, below). The subsets folder also contains a database (data.db, an hdf5 file) of information on the subsets of sites that were analysed, so that PartitionFinder can re-run analyses without re-calculating lots of results.

schemes folder contains detailed information on the schemes that were analysed during the analyses, each in a separate .txt file that is very like the best_scheme.txt file. For the greedy and clustering algorithms, this folder contains only the starting scheme and the best scheme that was found at each step of the algorithm. For the kmeans algorithm, it will just contain the start_scheme.txt and final_scheme.txt files, because we cannot save schemes along the way during the kmeans algorithm (read the paper to find out why). This folder will also contain a .csv file that summarises all of the schemes. If you analyse really huge datasets, you may want to turn off writing these files using the command line option -q (see below).

The table below shows the kinds of results returned by CIPRES Science Gateway:

Input File Names Sample File from a Test
phylip file part_finder_infile.phy
partition_finder.cfg partition_finder.cfg
Sample Output File Type File Name
zipped analysis directory
stdout file part_finder_stdout.txt
logfile part_finder_log.txt

If you use PartitionFinder2 here, please cite the appropriate works.

If you use any part of PartitionFinder2 in any published work please cite:
Lanfear, R., Frandsen, P. B., Wright, A. M., Senfeld, T., Calcott, B. (2016) PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Molecular biology and evolution. DOI:

If you use search = ‘rcluster’ or search = ‘hcluster’ please cite:
Lanfear, R., Calcott, B., Kainer, D., Mayer, C., & Stamatakis, A. (2014). Selecting optimal partitioning schemes for phylogenomic datasets. BMC evolutionary biology, 14 (1), 82.

If you use search = ‘kmeans’ please cite:
Frandsen, P. B., Calcott, B., Mayer, C., & Lanfear, R. (2015). Automatic selection of partitioning schemes for phylogenetic analyses using iterative k-means clustering of site rates. BMC Evolutionary Biology, 15 (1), 1

If there is a tool or a feature you need, please let us know.

hummingbird in flight

Get 1000 Hours free

On the UCSD Supercomputer

Start Your Trial