Distance programs

Programs designed and (except for dist.pl) implemented M. Goeker for conversion between (host and associate) character and distance data and related tasks. If you use any of the programs in a publication, please cite this web page. Each of the zip files with software contains three executables (for Linux, MacOS10, and Windows).

Data and distance formats

Examples for the formats used by the programs are provided. As a rule, the programs listed below are able to print their output in the popular NEXUS format used by programs such as PAUP* and MacClade. However, they are not able to read NEXUS-formatted files. The readme.txt file distributed together with the examples describes how to export data from PAUP* in a legible manner.

Computation of distance quality statistics

dist_stats: Computes distance quality scores for both entire matrices and indivdual taxa. Some of the scores have not been thoroughly tested; it is thus strongly recommended to rely on Delta values only, which have first been described by Holland et al.. However, almost all scores are described in our poster that also assesses the relative performance of the scores with two whole organelle genome datasets. Auch et al. and Goeker and Grimm have applied these scores to determine the best distance functions.
model2dist.nex: Batch file for PAUP* to to compute distance matrices along with likelihood scores. Whereas the scores can be input in Posada's famous modeltest program, delta values can be computed from the distance matrices with dist_stats as an alternative means to find the best model for phylogenetic analysis based on pairwise distances (e.g., Neighbour Joining). This technique was suggested by Holland et al. and was applied by Goeker et al.. However, in contrast to the application of modeltest, performance of that model selection method has so far not been examined by extensive simulation or empirical studies.
runds.sh: Shell script facilitating the analysis of the model2dist.nex output.

Computation of distances

eukdis: Computes distances from different kinds of character data with different formulae (default is Euclidean distances from quantitative characters). An older version has been used by Goeker and Grimm.
logdis: A distance conversion tool.
pbc: Computation of host distances from associate distances using two different methods, which are described in Goeker and Grimm. Have a look at gbk2fas for the extraction of host-associate information from Genbank flatfiles.
dist.pl: Perl script designed by M. Goeker and implemented by C. Oertel for computation of patristic (path-length) distances from trees and for averaging of distance matrices.
newick.tcl: Tcl script that extracts different types of information from trees. Can also be used to calculate patristic distances.

Conversion of characters

g2cef: Computation of host characters from associate characters using five different methods, which are described in Goeker and Grimm. The eukdis program (see above) can be used to compute distances from the converted character matrices. Have a look at gbk2fas for the extraction of host-associate information from Genbank flatfiles.
perma: Simple program to randomly permute columns of a tab-separated text file a specified number of times. Can be used in a UNIX pipeline to calculate permutational statistics.

Questions?

In case you have any questions regarding these programs, send an e-mail to support [at] goeker [dot] org. Please include the name of the program somewhere in the subject.