Genomics has not yet superseded Quantitative Genetics

Efficient tools for sequencing genomes were developed in the 1990s, and they have gotten about 1000x better since then. Everyone predicted that we would figure out what every gene did, and diagnose genomes based on those individual genes.

For the previous century, genetics had to cope with much more primitive methods. Biologists would have to study phenotypes, measure their heritability, and develop linear models for those unknown genes. It was like a car mechanic trying to fix a car without knowing how engines works.

The older methods could be called the infinitesimal model, or quantitative genetics. The new methods could be called genomics.

Here is a fascinating lecture on The Lost Evolutionary Synthesis.

He explains how the older infinitesimal model was spectacularly successful at breeder better chickens for farmers, and in solving lots of other problems. The new genomics has not had similar accomplishments.

I am still expecting genomics to catch up. A new research paper last month on Genome modelling and design across all domains of life with Evo 2 shows that maybe AI can finally understand genomes.

But as of today, the decades-old infinitesimal model has been much more productive.