PhyInC Logo: Phylogenetically Independent Contrasts Sequence Logo

PhyInC Logo (pronounced "Fink" for short) is a tool to take Fasta files and Newick tree files to create a phylogenetically conditioned sequence logo, using Felsensteins phylogenetic independent contrast (PIC).

Install from PyPI.org

pip3 install phyinc

It is always a good idea to use virtual environments, see below.

Install from a downloaded GitHub repository

For dependencies, start a virtual env as good practice:

> python3 -m venv .
> source venv/bin/activate
> pip3 install biopython weblogo matplotlib numpy

If you are on mac

> brew install ghostscript

To install the current package:

> pip3 install .

Usage

The basic usage is as follows.

> phyinc treefile fastafile

Since no outfile is given, the logo is output to fastafile.fa_logo.pdf. You can decide outputfile and format using the -o option:

> phyinc -o the_logo.png treefile fastafile

Examples

There is example data in the github repository, and there you can run this command:

> phyinc examples/synthetic_data/ex1_t1.tree examples/synthetic_data/ex1.fa

This should create a PDF named "ex1.fa_seqlogo.pdf" in the examples folder.

There are cases when a phylogeny is created on proteins but the logo is created for domains and this may cause protein accessions be something like ETA_STAAU but the domain accession is ETA_STAAU/96-110. You can then use the --coords option to ignore the domain coordinates when mapping the domain to a tree.

> phyinc --coords examples/PF000672.fa examples/PF000672.treefile

However, phyinc will report an error if there more than one domain per protein:

> phyinc --coords PS00027.fa PS00027.treefile
Error: 'ZFH2_DROME' is a protein appearing twice, probably because you have two 
domains from the same protein in the input. If so, you must submit a tree inferred
on the domain sequences, not on the proteins.

License

GPLv3

Authors

• Haolin Guo wrote the basic code as part of his BSc project.
• Kyle Tenn helped make the code into a Python/PyPI package.
• Lars Arvestad oversaw and helped with creating the final Python/PyPI package.