heibl / ips Goto Github PK

R Interfaces to Phylogenetic Software

R 100.00%

ips's Introduction

ips: interfaces to phylogenetic software

NOTE: ips contains most functions that were formerly included in the phyloch package plus some more recent functions. Those phyloch functions related to tree plotting have been moved to the viper package.

This is a bundle of functions that present interfaces to popular phylogenetic software for sequence alignment, masking of sequence alignments, and estimation of phylogenies and ancestral character states. In additions, there functions for reading, manipulating and writing phylogenetic data (multiple sequence alignments and phylogenetic trees).

Introduction

There are several functions for reading and writing DNA sequences in FASTA, PHYLIP, and NEXUS format: read.fas, read.phy, read.nex, write.fas, write.phy, and write.nex. Some functions are available for integrating BEAST with R. XML input files for BEAST can be generated with rbeauti. Two functions are designed to read TreeAnnotator output: read.beast will render an object of class phylo with additional node statistics appended as list elements. These additional node statistics will be lost be the subsequent use of ladderize or rotate (or similar functions that change the ordering of internal nodes).read.beast.table also parses the TreeAnnotator output, but returns a matrix of node statistics. This package itself does not implement techniques for phylogenetic analyses, but provides a series of wrappers for commonly used software packages. Sequence alignment can be done with the mafft and prank; cleaning of sequences with gblocks and aliscore. The function raxml and mrbayes are intended for phylogenetic tree search. Running mrbayes with argument run = FALSE can be used to create MrBayes-executable NEXUS files. Finally, wrappers are provided for Multistate in the BayesTraits package (see multistateML and multistateMCMC). Several plotting functions (HPDbars, clade.bars, box.clades, box.tips, tip.color, edge.color have been moved to the viper package.

Installation

ips is available via CRAN and can be installed:

install.packages("ips")

Or you can install the development version via GitHub with:

# install.packages("devtools")
devtools::install_github("heibl/ips")

ips's People

Contributors

Stargazers

Watchers

Forkers

andreas-bio richelbilderbeek brendanf

ips's Issues

wrapper for online version of mafft

Hi there,

I would like to use the mafft wrapper in my R workflow. But I can't get mafft software running on my windows pc, how hard I try to specify the exe file C:/Documenten/mafft-win/usr/bin/mafft. Would it also be possible to use the online version of mafft in the wrapper? https://mafft.cbrc.jp/alignment/server/

Best,
Berry

[request] mafft function

Are you still maintaining the CRAN version?

Can you please add the option to mafft to add options freely?
I really need the --unalignlevel --allowshift --keeplength options, but I have to fix() it every time.

Multiple functions are not exported

Multiple functions, such as read.beast, are not exported.

Implementing --reorder option in ips::mafft

Thank you for developing this function! It is helping me a lot!
Is it possible to implement the option "--reorder" so the output of the alignment is based on the alignments instead of in the input fasta file order?

**More specifically, I'd like to run:

mafft --reorder --adjustdirection --auto input

**I'm using the instruction:

ips::mafft(this.aligmment, method='auto', options = c("--adjustdirection","--reorder")

**But I get the error:

Error in match.arg(options, c("--adjustdirection", "--adjustdirectionaccurately")) :
'arg' must be of length 1

Thanks in advance!

read.beast.table removed?

Hello,

In R 4.0.2 (om MacOSX) I tried:

install.packages("ips")
library(ips)
?read.beast()
?read.beast.table()

Were these functions removed from ips?

Thanks

Fernando

Enable passing executable name/path for mrbayes(), and maybe others.

As per this question asked in Biostars, it would be nice to be able to pass the path to the executable in functions like mrbayes(), so that it can adapt to the installation idiosyncrasies. In particular, currently it seems problematic in Windows (which I haven't tested, but the OP on that Biostars question has).

Link to the babette R package suite in ips::rbeauti

Dear dr Heibl,

I feel ips is a good generalist R package, allowing to work with many phylogenetic tools. For example, the ips::rbeauti function creates a general BEAUti XML file, to be used by BEAST2.

I suggest to add a link in the ips::rbeauti documentation to a specialized R package suite [1] to work with BEAST2 called babette [2] (disclaimer: I wrote that package suite), so that people that want to do something more specialized than ips::rbeauti can better find this package.

If you agree, I volunteer to create a Pull Request with this suggestion, but if you prefer to do it yourself, would also be fine.

Do you agree this is a good idea?

Cheers, Richel Bilderbeek

[1] I use the word 'suite' as babette consists out of five packages
[2] Bilderbeek, Richèl JC, and Rampal S. Etienne. "babette: BEAUti 2, BEAST 2 and Tracer for R." Methods in Ecology and Evolution (2018). https://doi.org/10.1111/2041-210X.13032

Make package for conda-forge

@heibl I'm packaging ips for conda-forge. You don't have to do anything, but I'll add you as a recipe maintainer if you are interested.

In practice, there is very little to do to maintain R package recipes on conda-forge. There is a bot which will notice whenever you publish a new version on CRAN. It then updates the recipe accordingly and creates a pull request. If you are a maintainer, you'll get notified when that happens and the results of automated testing, and can take whatever actions are necessary (typically accept the pull request). Of course maintainers also get notifications if someone reports issues with the package on conda-forge.