Hybrid detection with HyDe

We can use HyDe to run a series of hytpothesis tests among rooted triplets for signal of hybridization. We will be following the tutorial here. HyDe is available in Python. From the terminal you can open Python with python3 or python (but this may open an older installation). We work on the code folder.

Data Preparation for running HyDe

HyDe needs the sequence data in Phylip format, however, our sequences are in fasta format. Thus, we will need to convert our sequences to Phylip format. We can do this in Python by using the BioPython module. If BioPython is not installed you can run the command pip3 install BioPython in the terminal.

from Bio import SeqIO

#read in the fasta format sequence
data_path = "../data/Wheat_Relative_History_Data_Glemin_et_al/"
concat_file = "triticeae_allindividuals_OneCopyGenes.fasta"
records = SeqIO.parse(data_path+concat_file, "fasta")

#convert the output to phylip
count = SeqIO.write(records, "../results/10-triticeae_allindividuals_OneCopyGenes.phylip", "phylip-relaxed")

Additionally we need a mapping file that maps individuals to their respective species. When performing our coalescent analysis, we created the mapping file 07-species-mapping.txt which is in the exact format that we need.

Running HyDe on the full concatenation

Now that we have all the components from the terminal we can run run_hyde.py. Note that this file was downloaded and set as executable in the PATH when we installed HyDe. You can check you have it by simply typing run_hyde.py anywhere in the terminal.

For this command, you need to be in the code folder:

run_hyde.py -i ../results/10-triticeae_allindividuals_OneCopyGenes.phylip -m ../results/07-species_mapping.txt -o H_vulgare -n 47 -t 17 -s 11354214 --prefix 10-hyde

[Josh, Figure out what is going on with run_hyde.py on your computer]

Note that in order to read the data, we needed to provide a little extra information. Specifically, we needed to provide the outgroup (H_vulgare), number of individuals sampled (47), number of species (17), and number of sites (11354214).

This command took like an hour to run.

We want to move the results to the results folder:

mv 10-hyde-out-filtered.txt ../results
mv 10-hyde-out.txt ../results

Running HyDe on 10MB windows

Similarly to how we ran RAxML on each of the 10MB windows, we can use a bash script to run HyDe on each window. However, we first need to convert all the windows to the appropriate format.

In code, we open python:

import os
from Bio import SeqIO

window_phylip_path = "../results/10concatenation10Mb_OneCopy-phylip/" 
os.makedirs(window_phylip_path,exist_ok=True)
window_path = "../data/Wheat_Relative_History_Data_Glemin_et_al/Concatenation10Mb_OneCopyGenes/"
all_windows = os.listdir(window_path)

for window in all_windows :
	records = SeqIO.parse(window_path+window, "fasta")
	count = SeqIO.write(records, window_phylip_path+window, "phylip-relaxed")

[Note the above code did not work for me, so alternative code in 10-fasta2phylip.py run from the terminal: python 10-fasta2phylip.py]

In the paper, the authors run HyDe on all chromosomes, but that turns out to be too time-consuming, so we are only going to run on chromosom 3 (to reproduce Figure 3b).

I am going to create another folder 10concatenation10Mb_OneCopy-phylip-ch3 and I will move all files corresponding to this chromosome there:

In results:

mkdir 10concatenation10Mb_OneCopy-phylip-ch3
mv 10concatenation10Mb_OneCopy-phylip/*Chrom_3* 10concatenation10Mb_OneCopy-phylip-ch3

Now we can run HyDe on each window with the bash script 10-hyde-10mb.sh:

In code:

./10-hyde-10mb.sh