I've been searching around to find some good resources on metagenomic shotgun analysis but I can't really find any. I've read the biostars handbook section but it mostly covers 16S analysis. I'm specifically interested in long read (a la Nanopore) metagenomic sequence assembly.
I'm a newbie to this field, and from what I can tell, a generic assembly pipeline would look something like FastQC -> CANU -> some sort of unsupervised binning software -> RACON and then on to classifying and such.
In theory my goal is, I want to be able to go out into the woods with a minION, sequence some random soil bacteria, and be able to tell what bacteria it is, what it does, and also be able to identify any novel bacterial genomes present.
Am I on the right track here and can anyone point me in the direction of good resources, or have advice?
•
written
by
jimpinney •
2 answers
Sounds interesting.
This paper might be useful:
https://www.nature.com/articles/s41598-020-70491-3
- Assembly - tools already mentioned, + metaflye
- Polishing
- Gene finding - prokka etc. Perhaps Maker ?
- interproscan for protein sequence analysis. Also diamond
- blastn vs genbank for finding novel contigs without hits
This resource cover all the steps of assembly and annotation:
https://github.com/parfreylab/metagenome_assembly_pipeline
There are links to software tools and recipes for individual steps.
Log in to answer this question.
-
Taxonomic classification of shotgun metagenomic data of rhizospheric soil (bacteria + fungi)
written by sumah.1si21bt017 •
-
shotgun metagenomics degs analysis
written by sooni •
-
Racon empty output (non-memory issue)
written by lajordac •
-
Metagenomic Binning: before or after denovo assembly?
written by gatheringdusthere •
-
A question regarding the differential coverage binning process in a metagenomic workflow
written by Rui •
-
N50 for de novo assembly result is too short
written by nisrinalulu •
-
Assembly quality check
written by tothepoint
-
Good coverage value for de novo assembly with NanoPore Reads?
written by marianaer98 •
-
Shotgun metagenomics: Why bother with de novo assembly?
written by BioinformaticsLad
-
Metagenomic Mappers & Benchmark Datasets?
written by jnowacki
my last take on this was a workshop discussion 2/3 years ago, when the error rate made it very experimental and a bitchallenging. I guess that still holds true, but this is what it makes exciting, too ;-)
I would fill the gap of the unsupervised binning software with metabat2, classification with Kajiu or Kraken2, but in general I would start with a pipeline close to what you sketched out.