News

Applications

Scaling Biology: From Theory to Practice

Oct 20, 2025

The Baker Lab recently published “Accelerating Protein Design by Scaling Experimental Characterization." Their results echo what we see daily at SynPlexity: breakthroughs in AI driven biology do not come from better models alone but from richer and larger experimental datasets.

Their results echo what we see daily at SynPlexity: breakthroughs in AI driven biology do not come from better models alone but from richer and larger experimental datasets.

This week, we had our first look at data from a 12,000 gene build using our multiplexed gene synthesis platform, DropSynth. Shallow sequencing showed performance well beyond expectations, a clear signal that large scale, high diversity synthetic gene libraries are on the road to becoming routine.

Scale matters because discovery is a statistical process. Every additional construct adds signal, depth, and diversity to the data that drive learning. At 12K, we are no longer sampling isolated points in sequence space. We are mapping landscapes. This opens the door for Broad Mutational Scanning and directed evolution libraries at a level that was previously inaccessible, enabling direct measurement of sequence function relationships across thousands of genes.

The Baker Lab’s work showcases how high throughput screening (HTS) transforms protein design. Our goal is to match that scale on the synthesis side, providing pooled DNA libraries that keep pace with modern screening capacity. Hitting 12K constructs in a single run demonstrates that SynPlexity can now scale to meet emerging demand cost effectively and efficiently. Back of the envelope calculations show roughly an eighty percent boost in process efficiency compared to prior generation methods.

The bottleneck is not biology. It is scale.

We are removing it.

Scaling innovation.