I have Raw data sequenced on Mi-seq NGS platform, the prime objective is to analyze the diversity of the given sequences. what could be the various pipelines and methods to distinguish the variable diversity ?
Hi Ahmad,
There are plenty of methods and pipelines and each of them have pros and cons. There is still no such software/pipeline suggested as a "gold standard" because the algorithms, reference databases etc. are sill changing and being enhanced.
However there are pipelines that are mostly used and recommend. The best way is not to use one pipeline, but to choose the most reliable and efficient algorithms/software at each analysis step, and currently I can suggest:
1) Removing primers, quality check etc.: DADA2 pipeine (using R environment, both linux and windows available), alternatively using Qiime2 software (only linux), which has DADA2 built-in
2) Taxonomy assignment:
a) reference database: RDP, SILVA or GTDB. Be aware, that GreenGenes is very outdated and thus not longer recommended.
b) taxonomy assignment pipeline/algorithm: RDP Naive Bayesian cassifier (which is built in DADA2 package ) or idTAXA (DECIPHER package).
3) Then you can move to the diversity analysis. You can use phyloseq package (still in R) or again, Qiime2 software.
There is also mothur pipeline and others. You need to read some papers and go through tutorials which are available online and then decide what is the best solution for you.
Theoretically you should obtain the same (at least similar) results using any of suggested pipeline, but the practice tells that they can differ significantly.
For example most of the popular taxonomy classifiers are prone to the miss-classifications and over-classifications.
Hi!
If it is 16S or 18S library you can check Qiime2 pipeline, there are a lot of nice tutorials on their website.
If it's 16s sequencing, QIIME2 should be the tool of first choice. For shotgun metagenomic sequencing, it depends on the analysis goal, yet Kraken2 or Metaphlan are tools good for working with binning (taxonomy labeling each read) or profiling (abundance table per sample) respectively
Having any specific analysis to do would require digging deeper, preparing a pipeline yourself or working with experienced bioinformatician
Chapter MG-RAST, a Metagenomics Service for Analysis of Microbial Co...
MG_RAST is a good pipeline. That was the only one I used some years ago and was quite successful.
Best
After 4 years of its use, I can t tell you much about the details and could be outdated. The best thing to do first is to read the manual attached (quick check, I found the one from 2017) you can check whether a new version is now available.
Best
Another very powerful tool and really easy to use is the FROGS pipeline.
http://frogs.toulouse.inra.fr/
https://academic.oup.com/bioinformatics/article/34/8/1287/4708232?guestAccessKey=a3b4c7f6-d18e-43a0-971e-8f6422c67f56
- Badges
- Science method