Google DeepMind invited me to visualize AlphaGenome, a new AI model designed to help us better understand how the genome functions, particularly how DNA variants and mutations can influence biological behavior.
The model can analyze long stretches of DNA and generate predictions at the resolution of individual DNA letters. While avoiding any specific applications and keeping it intentionally non-distinct in terms of practical use cases, I emphasized scalability and openness for future scenarios.
The model can analyze long stretches of DNA and generate predictions at the resolution of individual DNA letters. While avoiding any specific applications and keeping it intentionally non-distinct in terms of practical use cases, I emphasized scalability and openness for future scenarios.
Genome as a Landscape -
I visualized the genome as an abstract terrain that is being mapped and processed. The visual direction takes the form of a topographic landscape representing DNA, with peaks and valleys symbolizing genetic variants. The mapping is expressed through references, peaks, and tracks, while subtle details illustrate transitions and connections between molecular layers. Expansive, exploratory, and centered on discovery.
