Metagenomics & Microbiome

The full workflow — QC, host removal, assembly vs. read-based profiling, functional annotation. This EBI course is the most systematic free resource available.

Binning assembles contigs into draft genomes without culturing. Knowing how to evaluate completeness and contamination (CheckM metrics) is the difference between a useful MAG and garbage.

Kraken2, MetaPhlAn, mOTUs, Bracken — each makes different trade-offs between sensitivity and precision. Understanding these trade-offs prevents over-confident taxonomic calls.

The standard for 16S/ITS amplicon data. The plugin architecture is complex — start with DADA2 for denoising and a classifier trained on SILVA for taxonomy.

Kraken2 classifies reads against a reference database at high speed. Bracken re-estimates species-level abundances. Use them together, not separately.

MEGAHIT for large datasets (faster, lower memory). metaSPAdes when you have deep coverage and care about longer contigs. Run both on a test dataset to compare.

Estimates completeness and contamination of MAGs using lineage-specific marker genes. A MAG below 50% completeness or above 10% contamination should not be reported as a genome.

Applying long-read sequencing to amplicon metagenomics — the specific challenges of Nanopore error rates and how clustering approaches handle them.

Viral metagenomics presents unique challenges — no universal marker gene, high divergence, fragmented assemblies. This talk maps the computational strategies.

The algorithmic foundation of modern metagenome assemblers. Understanding how assembly graphs work helps you interpret why your assembler behaves the way it does.