phynd the conflict in your alignments
The scripts here are relatively simple and easy to run. But in order to take full advantage of them, you will want to ensure that these are installed:
- iqtree - for running the tree analyses
- R and ggplot within R - for plotting
- bp - for conducting conflict analyses
The package is just one script, however, it relies on the three packages above. Iqtree just needs to be in your PATH (so that when you type iqtree in the terminal it runs). The same with R if you want to plot (also ensure the ggplot library is installed). bp is the only odd one. For this, you need to have go installed.
Use your package manager (sudo apt install golang-go or what is relevant) or download from the website. Then run go get github.com/FePhyFoFum/gophy and then go build github.com/FePhyFoFum/gophy/bp/bp.go. This will generate the bp program that you can put in your PATH (probably sudo cp bp /usr/local/bin).
To get phynd, you can clone the repository with git clone https://github.com/FePhyFoFum/phynd.git or just download the archive.
You can run a basic analysis with the command
./src/phynd.py -s SEQFILE -w 1000 -i 1000 -t 2
and
./src/phynd.py -s SEQFILE -w 1000 -i 1000 -p -t 2
if you want to plot.
In this example, there is a 3000 bp sequence with conflict in the last 1000 bp. We can run the example like
cd examples/t1files
../../src/phynd.py -s t1.fa -w 500 -i 500 -t 2 -p
This will plot the results (-p) use 2 threads (-t 2) and includes a 500 sliding window and interval.
The results from this are below. You can see the conflict in the last 2 segments corresponding to the last 1000 bp. Each row in the plot corresponds to a clade. So all but one of the clades conflicts. Also, all of the segments conflict in one clade. In other words, the analysis has detected conflict. You can explore the clades in the gzip file that is output along with the data. In this case t1.fa.interval_plotdata.details.gz. This will show the specific conflicts (you may need to unzip it with gunzip t1.fa.interval_plotdata.details.gz).
