Which Tool and Database for metagenomic classification at Species level with 16S
3
0
Entering edit mode
4 months ago
Aytaç • 0

I want to perform a metagenomic profiling at Species level. I had troubles with 16S data when I tried to analyse it with kraken2. I couldn't reach Species level when I used SILVA and RDP database. Greengenes did perform but it is an old respiratory so i didn't want to use that.

Is there any tool or database to solve this problem? I need a tool that performs Species level classification by using 16S data.

Thanks.

Taxonomy 16S Amplicon Metagenomic • 696 views
ADD COMMENT
1
Entering edit mode
4 months ago

I had some very good Illumina 16S results a few years back using Ampliseq - https://nf-co.re/ampliseq/2.12.0. It takes a bit to get running but was well worth it.

ADD COMMENT
0
Entering edit mode

If its giving species level thats all i wanted sir. Kraken2 is good and fast but as i said its not enough to determine species level when you use 16S database. It does if you do with shotgun database but i dont think it would be meaningful for bioinformatics.

ADD REPLY
0
Entering edit mode

i found some species level match but its very low number of match. Is it possible to happen ?

ADD REPLY
0
Entering edit mode
4 months ago
michael.ante ★ 4.0k

A few years ago, I had good results with pure Dada2 ( https://benjjneb.github.io/dada2/index.html ). It is part of the AmpliSeq workflow, which colindaven mentioned.

Using the workflow described in Dada2 helps you to understand your data and the overall process, which is streamlined in said nf-core workflow.

[edit] Dada2 runs well with the precompiled SILVA database. Kraken2's k-mer approach might be a bit to coarse resolving species identification on 16S-amplicons.

ADD COMMENT
0
Entering edit mode

Thank you very much. I will surely check the workflow.

ADD REPLY
0
Entering edit mode
4 months ago
hakimbazol ▴ 60

You may make the database by yourself by using NCBI refseq and train it based on your data processing like QIIME2. Since you will make it manually, it takes time but the result quite worth it since you try to use your own database.

ADD COMMENT
0
Entering edit mode

I tried but i dont have enough RAM to process...

ADD REPLY

Login before adding your answer.

Traffic: 2358 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6