Sequence error correction - LangilleLab/microbiome_helper GitHub Wiki
There are several methods available that correct errors in amplicon sequences. We have included some of these tools in our virtual box image, including: DADA2, Minimum Entropy Decomposition (MED) and IPED. DADA2 and MED (with oligotyping) are alternatives to OTU picking, which is the direction the microbiome field is moving. Below is a quick overview of how you can try these tools out on our virtual box. For details you should check out each tool's website linked to above. Also, if you use any of these tools be sure to cite them correctly.
DADA2
DADA2 takes in a directory of FASTQ files and models and corrects errors found within the reads. This approach allows exact sequence variants to be inferred, which is an alternative to OTU clustering. This is the approach that QIIME 2 (which is still in alpha phase) is moving towards. There is a tutorial available for DADA2 that you can walk through. You'll be able to run this in R (on the command-line, but you can install RStudio yourself) on our virtual box.
Besides the DADA2 website linked to above, you can check out the DADA2 paper for details: Callahan et al. 2016. DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods 13:581-583.
Minimum Entropy Decomposition (MED)
MED is a core part of the oligotyping pipeline, which is described in detail here. It takes in a FASTA file (so no quality info, unlike DADA2) and decomposes reads into nodes rather than OTUs.
You can check out the MED paper for details:
Eren AM, Morrison HG, Lescault PJ, Reveillaud J, Vineis JH, Sogin ML. 2015. Minimum entropy decomposition: Unsupervised oligotyping for sensitive partitioning of high-throughput marker gene sequences. ISME J 9:968–979.
If you just want to run MED and aren't interested in the rest of the Microbiome Helper resources then you might be interested in the Meren lab's virtual box. If not, you can try it out on our virtual box with these commands:
cd /home/shared/err_corr_examples/med_example
decompose sponge-1K.fa -E sponge-sample-mapping.txt -T
Note that the "-T" option above turns of multi-threading, which might be necessary if you only have 1 CPU enabled on the virtual box. If you have more CPUs to spare and you omit this option the command should run much faster.
Once this job completes you'll be able to take a look at the below HTML file in FireFox which summarizes the analyses:
sponge-1K-m0.10-A0-M0-d4/HTML-OUTPUT/index.html
IPED
IPED is an algorithm designed to correct sequencing errors specifically in paired-end Illumina MiSeq reads. These reads can then be used for OTU picking.
If you want to try IPED out on our virtual box then you can use the below commands.
Step 1: Create assembled FASTA and custom quality files required by IPED
(Note that IPED requires full paths to be specified for all arguments)
/usr/local/prg/IPED/IPED_V1.run _F /home/shared/err_corr_examples/IPED_example/sample.forward.fastq _R /home/shared/err_corr_examples/IPED_example/sample.reverse.fastq _o /home/shared/err_corr_examples/IPED_example/step1_out_
Step 2: Run error correction algorithm
Note that you'll need to replace "TMP_DIRECTORY" below with the large number which is the name of the folder in the Step 1 directory (e.g. for me it was "952567556015314", but it will be different each time you run it).
mkdir step2_out
/usr/local/prg/IPED/IPED_V1.run _f /home/shared/err_corr_examples/IPED_example/sample.fasta _n /home/shared/err_corr_examples/IPED_example/sample.names _c /home/shared/err_corr_examples/IPED_example/step1_out_IPED_Final/TMP_DIRECTORY/sample.forward.trim.contigs.fasta _q /home/shared/err_corr_examples/IPED_example/step1_out_IPED_Final/TMP_DIRECTORY/sample.forward.contigs.qual _o /home/shared/err_corr_examples/IPED_example/step2_out
You can check out the IPED site or the paper below for details:
M. Mysara, N. Leys, J. Raes, P. Monsieurs. 2016. IPED: A highly efficient denoising tool for Illumina MiSeq Paired-end 16S rRNA gene amplicon sequencing data. BMC Bioinformatics 17:192.