Skip to content

Latest commit

 

History

History
147 lines (100 loc) · 6.24 KB

Qiime2_pipeline.md

File metadata and controls

147 lines (100 loc) · 6.24 KB
Raw

QIIME2

This is the readme file for the Qiime2 pipeline to use for the analysis of 16S amplicons for example! You can activate Qiime2 by the following command:

source activate qiime2

EMP format

The EMP (Earth Microbiome Project) format always contains the amplicon reads in two files: "reads.fq.gz" file and the "barcodes.fq.gz" file. We can get these two files separately from a single fastq file from any Illumina or Ion-Torrent machine by using the commnad from Qiime1 "extract_barcodes.py".

extract_barcodes.py -f IonTorrent_Run2.fastq.corrected -l 8 -m ../Asg_090217_sam_map.txt

Here the -l represents the length of the barcode and -m the mapping file for the samples.

Import and demultiplex sequences

First, we import the reads and barcodes into an artifact

qiime tools import --type EMPSingleEndSequences --input-path Bar_a_seq --output-path Amp170209.qza

Then, we demultiplex the reads into samples using the mapping file.

qiime demux emp-single --i-seqs Amp170209.qza --m-barcodes-file ../Asg_090217_sam_map.txt --m-barcodes-category BarcodeSequence --o-per-sample-sequences Amp170209.demux.qza

Summarizing samples

Here, we summarize the demultiplexed samples and their individual sequence counts and also to look at the sequence quality plot to choose parameters for trimming in the following steps.

qiime demux summarize --i-data Amp170209.demux.qza --o-visualization Amp170209.demux.qzv

Then, we can visialize the summary file in a web browser by:

qiime tools view Amp170209.demux.qzv

DADA2 analysis

After the demultiplexing step, the quality control, 'ASV' selection, chimera checking and many other qulaity steps are dobe by the pipleline DADA2 along with extracting the representative sequnces and building a table for the representatives. All of this is done by just one command:

qiime dada2 denoise-single --i-demultiplexed-seqs Amp170209.demux.qza --p-trim-left 15 --p-trunc-len 270 --p-n-threads 5 --o-representative-sequences Amp170209.rep-seqs-dada2.qza --o-table Amp170209.table-dada2.qza

Note that the --p-trim-left 15 is to remove the forward primer and --p-trunc-len is to remove the reverse primer and also based on how the quality of the sequences looked like from the previous step. Amp170209.rep-seqs-dada2.qza contains the representative sequences and Amp170209.table-dada2.qza contains their respective abundance.

Visualizing files

We can visialize the different characteristics of the reprentstavive sequences and the table by the following commands:

qiime feature-table tabulate-seqs --i-data Amp170209.rep-seqs-dada2.qza --o-visualization Amp170209.rep-seqs-dada2.qzv

qiime feature-table summarize --i-table Amp170209.table-dada2.qza --o-visualization Amp170209.table-dada2.qzv --m-sample-metadata-file Comb_4ITruns_map.txt

qiime tools view Amp170209.table-dada2.qzv
qiime tools view Amp170209.rep-seqs-dada2.qzv

Merging different runs

In case there are many sequencing runs using the same barcodes for different samples, the tables and the representative sequences can be mereged together.

Note: The parameters used above should be the same for all the runs

qiime feature-table merge-seq-data --i-data1 Amp161122.rep-seqs-dada2.qza --i-data2 Amp170531.rep-seqs-dada2.qza --o-merged-data merged_seq.qza

qiime feature-table merge --i-table1 Amp170531.table-dada2.qza --i-table2 Amp170602.table-dada2.qza --o-merged-table merged_table2.qza

Note: it is also to be noted that only two features can be merged togther once!

Taxonomic classifications

Now, we can classify the representative sequences to their respective taxonomic unit using the already existing reference sequences such as the SILVA database and CREST and so on.

qiime feature-classifier classify-consensus-vsearch --i-query All_Asg_16S_rep_seqs.qza --i-reference-reads ~/Files/Database/Qiime2/Silva_128_Q2/16S_SILVA128_99_otus.qza --i-reference-taxonomy ~/Files/Database/Qiime2/Silva_128_Q2/16S_SILVA128_99_taxa.qza --p-threads 5 --output-dir Silva_classified

qiime feature-classifier classify-consensus-vsearch --i-query All_Asg_16S_rep_seqs.qza --i-reference-reads ~/Files/Database/Qiime2/Crest_Q2/Crest_97_otus.qza --i-reference-taxonomy ~/Files/Database/Qiime2/Crest_Q2/Crest_97_taxa.qza --p-threads 5 --output-dir Crest_classified

Bar plots

The Qiime barplots can be plotted using the following commnad:

qiime taxa barplot --i-table ../All_Asg_16S_table.qza --i-taxonomy classification.qza --m-metadata-file ../Comb_4ITruns_map.txt --o-visualization silva_taxa-bar-plots.qzv

Exporting classifications and table

qiime tools export classification.qza --output-dir .
qiime tools export All_Asg_16S_table.qza --output-dir .

The classfication table has a name taxonomy.tsv as an output with taxonomy for each ASV and the table out put is in biom format!

Krona plots

Here, we would combine the taxonomy classifications and the abundance table from the biom table to make Krona plots

biom convert -i All_Asg-table.biom -o All_Asg_asv_table.tab --to-tsv
krona_qiime.py ../taxonomy.tsv ../../All_Asg_asv_table.tab

The above biom command will create a normal ASV abundance table. then, we combine the the taxonomic classification of the ASV to their abundance using my own krona_qiime.py command. This will create text files for each sample in the analysis. then we would combine the text files to make the krona plots.

ktImportText text_files/Str* text_files/Sec* -o Piran_silva.html