Overview

What is metagenomics?

Metagenomics is the study of the structure and function of sequences isolated and analysed from all the organisms in a bulk sample. Metagenomics is often used to study a specific community of microorganisms, such as those residing on human skin, in the soil or in a water sample.

Metagenomics usually refers to microorganism samples, whereas environmental DNA (eDNA), while using overlapping tools and analysis, refers to other groups of organisms (such as metazoans).

An example - a gut content analysis examining the community structure of bacteria (microbiome) via 16S amplicon sequencing would typically be referred to as a metagenome study. Whereas if the assessment of the gut content was instead exploring what the animal’s diet was (what plants they have eaten, for example), using another amplicon marker (e.g. Cytochrome b) would be an eDNA study.

Amplicon vs shotgun (whole genome) sequencing

While whole genome sequencing provides a comprehensive view of all the genetic variations within a sample, amplicon sequencing focuses on sequencing specific genomic regions (like the 16s rRNA gene). This targeted approach makes amplicon sequencing more cost-effective than whole genome sequencing.

Shotgun metagenomic sequencing, unlike 16S rRNA sequencing, can read all genomic DNA in a specimen rather than just one portion of a particular gene. Shotgun sequencing can simultaneously identify and profile bacteria, fungi, viruses, and a variety of other microorganisms, which is useful for microbiome research.

Pros and cons of amplicon vs whole genome sequencing:

Analysis of the microbiome: Advantages of whole genome shotgun versus 16S amplicon sequencing

Full length amplicon vs hypervariable regions

Typically

https://nf-co.re/ampliseq/2.9.0/docs/usage#taxonomic-classification

https://data.eresearchqut.net/paulw/public/mahsa_manuscript2/index.html

ASV vs OTU

Taxonomic assignments in nfcore/ampliseq are based on Amplicon sequence variants (ASV), inferred using the DADA2 software package by matching the sample sequences to the SILVA ribosomal RNA sequence database.

DADA2 infers sample sequences exactly and resolves differences of as little as 1 nucleotide.

SILVA provides comprehensive, quality checked and regularly updated datasets of aligned small (16S/18S, SSU) and large subunit (23S/28S, LSU) ribosomal RNA (rRNA) sequences for all three domains of life (Bacteria, Archaea and Eukarya).

Traditionally Operational Taxonomic Units (OTUs) have been used in 16S ampicon studies. More recently ASVs have been used, due their improved accuracy in identifying taxa, particularly genus and species. Basically, OTUs utilise a similarity clustering method to identify taxa, whereas ASV are generated by quantifying exact sequence matches to an amplicon database (e.g. Silva or Greengenes) and then statistically adjusting this using confidence thresholds.

The OTU method typically can identify 97% similarity (with any accuracy) whereas the ASV method can identify even single base-pair differences. This enables a finer resolution of taxa down to the genus and species level. Note that there is increasing ‘fuzziness’ toward the lower taxonomic levels, as the diversity within some taxa is greater than the diversity between this and other taxa (in other words, even with ASV, not all taxa can be resolved to lower taxonomic levels and this highly depedant on the taxonomic group involved).

https://www.zymoresearch.com/blogs/blog/microbiome-informatics-otu-vs-asv

Illumina vs Nanopore sequencing technologies

This overview is based on this excellent review article:

https://www.nature.com/articles/s41587-021-01108-x

Nanopore is part of the ‘third generation sequencing’ suite of sequencing technologies, producing longer (~500bp - 100,000bp), fewer reads than the more commonly used ‘second generation’ of sequencing, most typically represented by Illumina fixed-length, short read technology (~50bp - 400bp).

Whereas 2nd generation technology uses massively parallel sequencing - i.e. simultaneous sequencing millions of small DNA fragments annealed to a flow cell - 3rd generation sequencing includes several competing technologies, which differ substantially in terms of underlying technology. The two main companies are Pacific Biosystems ('PacBio') and Oxford Nanopore Technology ('ONT' or ‘Nanopore’).

An overview of 2nd and 3rd gen sequencing technologies can be seen here: https://www.sciencedirect.com/science/article/pii/S0198885921000628

It’s important to note that a significant difference between 2nd and 3rd generation technology is accuracy. The error rate (i.e. the number of bases with low sequencing quality scores) of 3nd gen has been considerably higher than 2nd gen, with typically ~0.1% error rate for Illumina sequences and >5% error rate for Nanopore. The Nanopore error rate has improved dramatically in recent years though, but still is considerably lower than Illumina. This higher error rate can cause issues, such as in metagenomics when identifying species that differ by a small number of base pairs.

Functionally, the longer 3rd gen reads can counter the higher error rates through increased number of potential base matches. For 16S rRNA sequencing, short read Illumina sequences typically cover two 16S hypervariable regions, whereas Nanopore sequences the full 1.5 kilobase 16S sequences, which includes all nine hypervariable regions.

This 2023 paper compared Illumina and Nanopore shotgun sequencing for identifying bacteria strains with little genomic variation between them. Both Illumina and Nanopore were able to correctly identify the bacteria strains, despite the higher error rate of the Nanopore sequences.

Reference paper

Data used in this workshop is from a paper that compared Illumina and Nanopore 16S datasets.

https://www.mdpi.com/2073-4425/11/9/1105

2.8. Sequence Data Availability

The Illumina and nanopore sequence datasets of the nose swab samples, generated and analysed in the current study, are available in the European Nucleotide Archive (ENA) under accession number PRJEB28612

https://www.ebi.ac.uk/ena/browser/view/PRJEB28612

Why Nextflow?

What is Nextflow was covered in session 1 of these workshops (Installing Nextflow ).

“scalable and reproducible scientific workflows using software containers.” https://www.nextflow.io/

Bioinformatics workflows are complicated, and are becoming more complicated. Nextflow enables managed, reproducible curated bioinformatics workflows to be used by non-bioinformatician researchers.

Why do we use Nextflow?

1. Complexity: Analysis workflows are becoming more complex, with more steps (but producing more accurate, publishable results).

• 100’s of published workflows

• Curated, multi-tool analyses

• Optimised and improved over multiple versions

• Detailed output with results, tables, figures, data files

2. Curated: Managed workflows are typically tested and assembled by experts in the field. Often are improved over multiple versions.

• Nfcore: central repository for Nextflow pipelines: https://nf-co.re/pipelines

• Nanopore has developed Nextflow workflows via epi2me: https://labs.epi2me.io/wfindex/

3. Reproducibility:  Most published studies can’t be reproduced. The ‘reproducibility crisis’. Nexflow has version control, for both the workflow, and the multitude of software tools within.

The Nextflow workflows we’ll be running today are:

nfcore/ampliseq for the Illumina data: https://nf-co.re/ampliseq/2.9.0

wf-metagenomics for the Nanopore data: https://github.com/epi2me-labs/wf-metagenomics