Taxonomy assignment using Silva 138

Hi Nicholas,

I apologize for my lack of clarity. I meant that the dereplication step was successful, but after dereplication, training the classifier still failed and the job was killed, likely due to memory.

However, once I got my own Silva data using the RESCRIPt pipeline (following all the way through steps 1a through 1e). After dereplication I applied an additional taxa filter-seq step as recommended here before training the classifier (feature-classifier fit-classifier-naive-bayes) and taxonomy assignment (feature-classifier classify-sklearn). It was after this, I was able to finally carry out these last 2 memory intensive steps, but as mentioned. these last 2 steps only took 10-15 minutes to finish... so maybe something went wrong after all.

###dereplication of Silva full length sequences:
qiime rescript dereplicate
--i-sequences silva-138.1-ssu-nr99-seqs-filt.qza
--i-taxa silva-138.1-ssu-nr99-tax.qza
--p-mode 'uniq'
--o-dereplicated-sequences silva-138.1-ssu-nr99-seqs-derep-uniq.qza
--o-dereplicated-taxa silva-138.1-ssu-nr99-tax-derep-uniq.qza

###making amplicon specific classifier
qiime feature-classifier extract-reads \ ###16S
--i-sequences silva-138.1-ssu-nr99-seqs-derep-uniq.qza
--p-f-primer GTGYCAGCMGCCGCGGTAA \ #515F
--p-r-primer CCGYCAATTYMTTTRAGTTT \ ###926R
--p-n-jobs 2
--p-read-orientation 'forward'
--o-reads silva-138.1-ssu-nr99-seqs-v4v5.qza

###dereplication of amplicon specific region
qiime rescript dereplicate \ ###16S
--i-sequences silva-138.1-ssu-nr99-seqs-v4v5.qza
--i-taxa silva-138.1-ssu-nr99-tax-derep-uniq.qza
--p-mode 'uniq'
--o-dereplicated-sequences silva-138.1-ssu-nr99-seqs-v4v5_derep.qza
--o-dereplicated-taxa silva-138.1-ssu-nr99-tax-v4v5_derep.qza

###filter seq step
qiime taxa filter-seqs
--i-sequences /home/Rocks/outputs/rescript/silva-138.1-ssu-nr99-seqs-v4v5_derep.qza
--i-taxonomy /home/Rocks/outputs/rescript/silva-138.1-ssu-nr99-tax-v4v5_derep.qza
--p-exclude Eukaryota,Mitochondria,Chloroplast,Unassigned
--o-filtered-sequences 16S_v4v5_derep_filt.qza

###train amplicon specific classifier
qiime feature-classifier fit-classifier-naive-bayes
--i-reference-reads /home/Rocks/outputs/rescript/16S_v4v5_derep_filt.qza
--i-reference-taxonomy /home/Rocks/outputs/rescript/silva-138.1-ssu-nr99-tax-v4v5_derep.qza
--o-classifier classifier_16S_v4v5.qza --verbose &> classifier16S_training.log & disown

###assign taxonomy
qiime feature-classifier classify-sklearn
--i-classifier /home/Rocks/outputs/classifier_16S_v4v5.qza
--i-reads /home/Rocks/outputs/qza_intermediates/rocks16S_rep_seqs.qza
--o-classification rocks16S_taxonomy.qza
--p-reads-per-batch 10000
--p-n-jobs 1
--verbose &> classify16S_verbose.log & disown

Thank you.