Require Pre-trained Silva Classifier (V3-V4)

General Discussion
1

I am looking for a naive-bayes classifier based on Silva reference database for V3-V4 region, 341F, 805R, generated with scikit-learn version 0.24.1.

Thanks in advance

2

Hello,

it is a custom classifier, you can train it using RESCRIPt, detailed tutorial is on the forum here: Processing, filtering, and evaluating the SILVA database (and other reference sequence data) with RESCRIPt.

Cheers,
Valentyn

3 Likes
3

Dear Priti, I have made databases based on Silva 132 database v3-v4, primers, based on scikit-learn 0.24.1. You need the ref-taxonomy, ref-seqs.qza and classifier.qza files, right?

qiime tools import --type 'FeatureData[Taxonomy]'
--input-format HeaderlessTSVTaxonomyFormat
--input-path consensus_taxonomy_7_levels.txt
--output-path ref-taxonomy.qza

qiime feature-classifier extract-reads
--i-sequences 99_otus_16S.qza
--p-f-primer ACTCCTACGGGAGGCAGCAG
--p-r-primer GGACTACHVGGGTWTCTAAT
--p-min-length 100
--p-max-length 500
--o-reads ref-seqsV3V4.qza

and

qiime feature-classifier fit-classifier-naive-bayes
--i-reference-reads ref-seqsV3V4.qza
--i-reference-taxonomy ref-taxonomy.qza
--o-classifier classifier.qza

I have not yet been able to check their quality, so I guess that I best send them (how?). I can publish them too (again zero guarantees). I also have a 138 rescript classifier, but I see now that it may be incorrect, possibly based on the 132 ref-taxonomy.qza.

4

RESCRIPt can generate a database from various versions of SILVA, including 128, 132, 138, and 138.1

SILVA version 138 has some known taxonomy issues, which is why SILVA released version 138.1 as a patch. This is perhaps the issue you spotted with your classifier? Pls open a separate topic if you want to report an issue.

2 Likes
5

@Nicholas_Bokulich No it was my mistake. I am re-running the fit-classifier-naive-bayes again, with the (hopefully) correct taxonomy. Rescript 138 database data come from: Data resources — QIIME 2 2021.11.0 documentation under Silva (16S/18S rRNA) but I have no chance to check the coming weeks, nor to look at making a database on 138.1.

1 Like
6

Thank you for the details

7

Thank you @jack2017 I prepared the SILVA database.

8

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.