Viromics2026: Glossary

Viromics is the study of viruses, and in our case bacteriophages, using next-generation sequencing technologies

Bacteriophages are diverse and ubiquitous across all biomes

Bacteriophages have large implications on their environment including the human gut

your laptop can connect to Draco

You now have a personal workspace with scripts and sequencing data, which we’ll use for the rest of the course

pwd, ls, and cd are the three commands you’ll use constantly to orient yourself

ls -ltrh sorts by modification time, oldest to newest, with human-readable sizes

The up/down arrow keys recall previous commands

Relative paths depend on where you currently are; absolute paths always start from /

mkdir, cp, mv, and rm manage files and directories. rm has no undo, use carefully

The most common file formats are FASTA (nucl. and amino acid), FASTQ, SAM/BAM, VCF, GFF/GTF, BED, and plain TSV metadata tables

A valid FASTQ record always has 4 lines, with the sequence and quality lines the same length — mismatches signal a corrupted file

cat, head, tail, and less view plain-text files. zcat, zless, and zgrep do the same for gzip-compressed files

Match your inspection tool to the file

Draco has a login node for light tasks and many compute nodes for resource-heavy hobs

Because Draco is shared, the Slurm job scheduler decides which compute node your work runs on and when

An sbatch script describes both the resources you need and the commands to run

.out holds normal output

.err holds error messages — check .err first when a job fails

Sequence assembly can be used to assemble genomes from reads

Metagenome assembly generally yields shorter contigs than genome assembly

Flye can be used to assemble long and noisy nanopore reads from metagenomic samples.

Samples can be assembled individually and combined in a cross-assembly

vClust can be used to assess the diversity of sequences in your assembly

Bandage can visualize the de-Bruijn graph

JBrowse2 can visualize genomic data like alignments and coverage

geNomad combines two independent branches: an alignment-free model that reads the nucleotide sequence directly, and a gene-based model that uses marker protein profiles

Going through a paper together at the board is fun. Jeroen is amazing at drawing :)

Filtering contigs by completeness and contamination is crucial to obtain an informative dataset

Tools like geNomad classify your contigs, enabling you to understand your samples

No wet-lab or dry-lab technique is perfect. Filtering non-viral contigs from your data improves its quality, helping you obtain better results

Genome annotation gives meaning to genomic sequences

ORFs can be predicted from start and stop codons in the genomic sequences

Phages have different genomic features than prokaryotes, which influences the design of algorithms

Gene calling predicts the coordinates, strand, and protein sequence of candidate genes

Phage genomes have distinctive features (e.g. dense/overlapping genes, atypical start codon usage) that differ from prokaryotic genomes, requiring specialized tools like Phanotate

Sequence-based annotation (e.g. Pharokka) assigns function by comparing predicted proteins against known viral genes

Sequence similarity alone often fails on divergent phage proteins.

Most genes on viral contigs remain unannotated due to rapid viral evolution, highly mosaic genomes, and how little of viral sequence space is characterized in reference databases

Using structure-based search or genomic context can be used to detect distant homologies, allowing for the annotation of more viral genes.

Viruses have multiple origins, so there is no universal marker gene

Viral taxonomy is based on many different methods

Gene-sharing networks and marker genes are popular methods for bacteriophage taxonomy

Marker genes such as the terminase large subunit (terL) can also be used to judge how related viruses are and in some cases classify the taxonomy as lower taxa ranks

RaFAH uses a random forest model to predict hosts to the genus level for phages

RaFAH returns a probability for each host genus it can predict.

PhageTransformer is a codon-aware genome language model that predicts a host from DNA sequence directly.