Training V3V4 classifier w/ rescript

[Sabrin (Abdelghany)]

@SoilRotifer

You can make your own 16S rRNA V3V4 region classifier using RESCRIPt . If you do not want to make the classifier yourself you can search the forum for others that have already made a V3V4 classifier. Many are willing to share.

In train my classifier using RESCRIPT the first step here Processing, filtering, and evaluating the SILVA database (and other reference sequence data) with RESCRIPt is to Getting SILVA data using this command
qiime rescript get-silva-data
--p-version '138'
--p-target 'SSURef_NR99'
--p-include-species-labels
--o-silva-sequences silva-138-ssu-nr99-rna-seqs.qza
--o-silva-taxonomy silva-138-ssu-nr99-tax.qza

but i get qiime: command not found when i type that in my terminal

So i went up to the Getting SILVA data: Hard Mode. and followed steps successfully until this command
Import the Taxonomy Rank file:

qiime tools import \
    --type 'FeatureData[SILVATaxonomy]' \
    --input-path tax_slv_ssu_138.txt \
    --output-path taxranks-silva-138-ssu-nr99.qza \

but it never give me output, what could be wrong here? Does this command needs too much time?

[SoilRotifer (Mike Robeson)]

Did you follow the RESCRIPt plugin install instructions? If not, follow the Github link (provided in the RESCRIPT plugin link I provided) and follow the "Option 2" install instructions.

[Sabrin (Abdelghany)]

Yes i followed the RESCRIPt plugin install instructions, option 2
with no problem and I am following until make our classifier for use on full-length SSU sequences. now as follows

time qiime feature-classifier fit-classifier-naive-bayes --i-reference-reads silva-138-ssu-nr99-seqs-derep-uniq
.qza --i-reference-taxonomy silva-138-ssu-nr99-tax-derep-uniq.qza --o-classifier silva-138-ssu-nr99-classifier.qza
Killed
Why I receive killed after several trials with this step please? I have 15.8 GB memory and it is always killed after reaching 13 GB of the memory!! What should i do so it is not killed every time please

For later steps, i received from sequencing center an email with primers used, message includes **illumina primer and primer, which one is the one intended to be used? I deal with v3/v4 16S

[SoilRotifer (Mike Robeson)]

@Sabrin

Why I receive killed after several trials with this step please? I have 15.8 GB memory and it is always killed after reaching 13 GB of the memory!! What should i do so it is not killed every time please

Often users require upwards of 32-64 GB of RAM to train the full-length sequence SILVA classifier. If you are going to use the full-length classifier, you can simply use the one we provide here. You might be able to construct the V3V4 region specific classifier with 16 GB RAM as it 'll be much smaller.

@Sabrin

For later steps, i received from sequencing center an email with primers used, message includes **illumina primer and primer, which one is the one intended to be used? I deal with v3/v4 16S

If you perform an internet search for "V3V4 primer sequence" you'll find that those listed in the Primer column are the primers you'd use as input into cutadapt.

[Sabrin (Abdelghany)]

Thank you so much.

So, succeeding with the amplicon-region specific classifier is enough for me so i will be following the rest of tutorial.

So, build our amplicon-region specific classifier is last step i need in the rescript tutorial , right.?
Sorry if my questions looks naive, i am just a beginner

[SoilRotifer (Mike Robeson)]

@Sabrin

So, build our amplicon-region specific classifier is last step i need in the rescript tutorial

Yep. Though keep in mind, that there is no one way to prepare a classifier. This is why we made RESCRIPt, so that you have as much control as possible to construct the classifier as you'd like.