Question about substitution of KEGG

Question

[dxh0222]

Hi, I’m currently working on functional annotation of ORFs predicted from assembled metagenomic contigs. Because the dataset contains a very large number of ORFs, it’s not practical to use KEGG’s online services such as BlastKOALA.

I also looked into eggNOG-mapper, but the database seems a bit outdated, and I’m not sure if it includes the most recent pathways or functional gene families. At the same time, a KEGG subscription is unfortunately beyond my budget.

So here’s the question I’ve been wrestling with: what are good, free alternatives to KEGG for large-scale functional annotation? Ideally, I’d like tools or databases that can handle millions of ORFs efficiently and still provide pathway-level insights.

Thank you for your guildance!

Answer 1

[Mensur Dlakic]

There is a tool that uses KEGG hidden Markov models and HMMer to scan individual proteins. It is not automated and may require some fiddling, but it will give you KEGG-like annotations.

https://github.com/takaram/kofam_scan

To just annotate all the proteins without necessarily grouping them into pathways:

https://github.com/oschwengers/bakta

An older automated annotation package:

https://github.com/tseemann/prokka

[dxh0222]

Thanks for your recommendation! I think kofam_scan is suitable for me.