This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Can i use orthofinder for small protein datasets and not full proteome?

Hi all,

I have predicted the effector proteins for a range of fungal genomes and i have them in fasta sequences. In order to compare them i need to group them somehow. I thought of using orthofinder but i dont know if its the correct approach or i will introduce any biases. My question is can i use orthofinder grouping (not the phylogenetic tree)? Does anyone has advice on a diferrent way to group them?

orthofinder effectors
• 2,042 views • link
updated by Emin Bursa • Emin Bursa • written by alexandrakortsi • alexandrakortsi

Do you want to compare them by sequence or by function?

For the later, you can use e.g. InterProScan, KoFamScan.

• 0 views • link
written by michael.ante michael.ante

2 answers

If you're just using the tool for clustering purposes, I think OrthoFinder might be a little clunky and would recommend using something like mmseqs2 easy-cluster instead. It's fast and one of it's intended purposes. But in theory you could use OrthoFinder just for clustering if you really want to stick to the same tool.

• 0 views • link
written by dthorbur dthorbur

You can use OrthoFinder just for grouping, and it’s a solid choice: run it on the full proteomes for each fungus, then afterward just look at which proteins in each orthogroup are predicted effectors—that gives you “effector families” across species with minimal bias. If you only run it on effector sequences, it still works but will exaggerate how many are species-specific. For extra resolution, you can later cluster the effectors themselves with an all-vs-all similarity + MCL or group them by shared domains (InterPro/Pfam), but OrthoFinder-on-full-proteomes + marking effectors is a very reasonable main strategy.

• 0 views • link
written by Emin Bursa • Emin Bursa

Log in to answer this question.

More posts like this