soedinglab / WIsH

Predict prokaryotic host for phage metagenomic sequences
GNU General Public License v3.0
52 stars 10 forks source link

How to get the p-values? #9

Closed ErnieHsieh closed 2 years ago

ErnieHsieh commented 2 years ago

Hi,

Thanks for developing the WIsH for phage-host prediction! I would like to kindly consult you a question when running the WIsH in my study.

First, I have run the default command to predict the hosts for my vOTU genomes. I used my taxonomy-assigned, binned bacterial MAGs (from my whole metagenomes) as the bacterial models separated by samples, and used my non-redundant vOTUs as target genomes.

Then run ‘wish -c predict -g ./phageContigsDir/ME_WMS_vOTUs_5kb_genomes -m ./modelDir/17-106 -r ./outputResultDir/17-106 -b 1’. As expected, I got the output but haven’t got the p-values for my prediction.

So next, I would like to obtain the p-value matrix.

I downloaded the reference-phage genomes from http://millardlab.org/bioinformatics/bacteriophage-genomes/phage-genomes-apr2022/

I used these downloaded reference phages as targets and used my binned bacterial MAGs as models. Then I ran the same command again. I assumed I will obtain those reference viruses/phages ‘infecting’ my bacterial bins. But in this case, should I also need to know those reference phages ‘NON-INFECTING’ my bacterial bins? Can I ask you how exactly I can do to obtain this?

I am new to use this tool so have not fully understand how I should/could run the null parameter section to get the p-values. Before getting the null-phage sets, how could I do for getting the non-infecting ones?

I would appreciate if you can elaborate on how I should do in my case to obtain the p-values please?

Many thanks in advance for any help and comment.

Kind regards,

Ernie