rcsbiotech/q2pipe

Microbiome pipeline optimised for completeness and algorithm diversity

Warning

I've stopped using QIIME2 entirely, so be warned if anything below doesn't work for newer (2019/06) QIIME2 versions.

q2pipe

QIIME2 16S wrapper Implemented for QIIME2 2018.8

Runs a complete OTU analysis on your sequence data based on a range of optimizable parameters, such as trimming and truncate length, annotation identity, and cluster parameters.

This is a side project on my Masters Degree, and i will be implementing features and user-friendliness as i go. Any questions or suggestions, you can easily reach me on the e-mail [email protected], with the subject "q2pipe: your subject"

What you must install before running:

• Conda package manager package manager that installs QIIME2 easily;
• QIIME2 installed via conda, the main analysis package;
• BIOM format for data wrangling;
• PICRUSt metagenome predictor;

What you must feed the program

• Analysis parameters, such as name, trimming and truncation values, max. expected errors, which algorithm to use for annotation, and such.
• A manifest text file for data;
• A sample metadata text file;
• Your paired-end sequences.

What does the wrapper do

1. Imports your data as a QIIME2 artifact;
2. Denoise it using the DADA2 pipeline;
3. Classifies your data using either BLAST or VSEARCH, against the latest Greengenes and SILVA database;
4. Run core diversity and basic phylogenetic analysis based on QIIME2
5. Predict the metagenomes with PICRUSt;
6. Reports all results as HTML, plain text and a usable BIOM-taxonomy file.
7. (WIP) Run our custom RScript with the main diversity indexes.

Suggested workflow

1. Turn on your qiime2 (usually with source activate qiime2-2018.8)
2. Create a manifest file, with the name manifest.csv, then put it inside the intel directory.
3. Run q2manifest to import your data as a QIIME2 artifact.
4. Open the demux-summary.html in a web browser, and pick parameters suitable to your sequences.
5. Write these parameters on intel/parameters.txt
6. Run q2pipe and then q2picrust.

And that's it.

Current schema

Currently in development

• Add user friendly description and help messages
• Error control for core wrapper (manifest, preproc, pipe and picrust)
  • Missing parameters;
  • Missing files;
  • Missing databases;

Roadmap before alpha

1. (done) Annotation with VSEARCH/BLAST over SILVA132 and the latest Greengenes.
2. (done) Core diversity analysis with QIIME2.
3. (WIP) Update to QIIME-2011-11.
4. Implement shi7 to test over parameters.
5. Merge-Denoise-Deblur optimized preprocess pipeline.
6. Automatic q2-feature-classifier.
7. Generate numerical ecology analysis based on metadata columns, on a custom RScript.
8. Implement for ITS regions.

---

References

1. QIIME2.
2. I can't thank enough Edgar at Drive5 (https://drive5.com/about.html) for many, many insights over the years.
3. DADA2 sequence denoise pipeline: Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016). DADA2: high-resolution sample inference from Illumina amplicon data. Nature methods, 13(7), 581.
4. VSEARCH for sequence clustering and feature classification: Rognes, T., Flouri, T., Nichols, B., Quince, C., & Mahé, F. (2016). VSEARCH: a versatile open source tool for metagenomics. PeerJ, 4, e2584.
5. BLAST also used for feature classification: Camacho, C., Coulouris, G., Avagyan, V., Ma, N., Papadopoulos, J., Bealer, K., & Madden, T. L. (2009). BLAST+: architecture and applications. BMC bioinformatics, 10(1), 421.
6. PICRUSt for metagenome prediction: Langille, M. G., Zaneveld, J., Caporaso, J. G., McDonald, D., Knights, D., Reyes, J. A., ... & Beiko, R. G. (2013). Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nature biotechnology, 31(9), 814.