Minimum length when filtering AMPs

[Sachintha Wijegunasekara]

Hi, 

Thank you for developing Macrel.

I am using Macrel and would like to know if there is a minimum length I should use for AA sequences when filtering the AMPs detected by Macrel.

I can see some AMPs with around 20 AA lengths for some of my samples.

Thank you,

Sachintha

[Luis Pedro Coelho]

Thanks for the question.

There is no minimum length. Macrel will not predict any very short sequences, but IIRC, at some point we fixed it so that it would even accept a 1 amino acid input without crashing (it will just predict it as non-AMP, obviously)

Best, LUis

--

You received this message because you are subscribed to the Google Groups "ampsphere-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ampsphere-use...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/ampsphere-users/442e50cc-68e0-40d3-901e-67301948ee50n%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Luis Pedro Coelho | Queensland University of Technology | https://luispedro.org

https://orcid.org/0000-0002-9280-7885

[Sachintha Wijegunasekara]

Thank you.

Additionally, I was considering the maximum lengths. The AMPSphere paper mentioned that the average lengths of AMPs are around 37 residues. However, I have some AMPs with longer sequences. I've used the nfcore/funscan pipeline to predict the AMPs with Macrel and several other tools. Since the maximum length of AMPs is around 50 residues, would it be best to exclude larger peptides, even if they're predicted as AMPS? 

Also, I compared my peptide sequences with AMPSphere peptides using mmseqs, which produced a table with a number of columns. Apart from the e-value column, is there any column I should consider when filtering? I'm having trouble locating the column descriptions. 

Finally, are there any additional steps I can take to ensure I have a high-quality set of AMP candidates?

Thank you,

Kind regards,

Sachintha

---

From: ampsphe...@googlegroups.com <ampsphe...@googlegroups.com> on behalf of Luis Pedro Coelho <lu...@luispedro.org>
Sent: Tuesday, June 24, 2025 10:50:26 AM
To: ampsphere-users <ampsphe...@googlegroups.com>
Subject: Re: Minimum length when filtering AMPs

 

You received this message because you are subscribed to a topic in the Google Groups "ampsphere-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ampsphere-users/tROy5MFUQWc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ampsphere-use...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ampsphere-users/8cd683f1-3f8c-4d86-aac3-55be352ea5d5%40app.fastmail.com.

[Luis Pedro Coelho]

We generally recommend that you filter sequences above 100aa for macrel as the tool has not been trained on those. Best, Luis