Genome Annotation

One-Sentence Definition

Genome annotation locates genes and other features on an assembled genome and assigns predicted functions to them.

Simple Explanation

After assembling the DNA, annotation writes the labels: “here is a gene, and it probably makes a β-lactamase.”

Detailed Scientific Explanation

Two layers:

  1. Structural — find CDS, rRNA, tRNA, ncRNA, CRISPR arrays (Prodigal, Barrnap, tRNAscan)
  2. Functional — assign names/ontologies by homology or profiles (BLAST/DIAMOND, HMMER + Pfam, KEGG, COG, GO)

Common bacterial pipelines: Prokka, Bakta, PGAP (NCBI).

Caveats:

  • Function transferred by homology is a hypothesis, not proof
  • “Hypothetical protein” is very common in non-model organisms
  • Annotation version changes downstream gene counts and pangenome results

Mechanism

Gene prediction models (Markov models trained on coding statistics) + alignment/profile search against curated reference sets.

Clinical Importance

Research Importance

Diagnostic Relevance

  • Structured input for clinical genomics reports

AMR Relevance

  • Correct CDS boundaries matter for detecting truncations (e.g., porin loss → resistance)

Active Recall Questions

  1. Structural vs functional annotation?
  2. Why is homology-based function a hypothesis?
  3. Which resistance mechanism shows up as a truncated gene?

Connections