Skip to main content
GemomicMicrobiology

Painting with bacterial genomes: BactoPainter turns genes into colour

By August 24, 2026No Comments

Painting with bacterial genomes

Built by Adrien Assié
Bioinformatics meets generative art.

E. coli K-12 MG1655 and H. pylori 26695 rendered on the same absolute biological colour scale. Every tile is one annotated coding sequence.

01 — THE IDEA

A genome is also a palette

The result sits between genomic overview, exploratory instrument, and generative art. It is not decoration placed over the data. The data determines the image.

02 — THREE SIGNALS, ONE COLOUR

RED

Codon Adaptation Index

Brighter red indicates codon usage better adapted to the genome—a possible signal of highly expressed genes.

GREEN

GC content

The proportion of guanine and cytosine carries a strong taxonomic and compositional signature.

BLUE

Karlin distance

An atypical dinucleotide signature can flag genes worth investigating for horizontal transfer.

03 — SAME GENOME. NEW PERSPECTIVES.

Three H. pylori strains, one visual language

Shared fields reflect the species’ broadly similar composition. Brighter interruptions reveal local changes in GC content, codon adaptation, or dinucleotide signature.

Important: positions are normalized, not homolog-aligned. A matching band is a hypothesis to investigate—not proof of the same gene.

04 — THE WORKFLOW

From sequence file to colour field

01

Read an annotated GenBank file, or FASTA with GFF annotation.

02

Locate OriC from annotation or cumulative GC skew.

03

Calculate CAI, GC content, and Karlin distance for every CDS.

04

Export biological values and RGB channels; render grid, barcode, spiral, or circle.

python main.py genome.gbk –color-mode absolute -o genes.tsv

A picture that leads back to the sequence

Explore BactoPainter on GitHub