Generative Biology

A fundamental shift in therapeutic development, driven by machine intelligence.

We asked the question.

What if… We could generate novel protein therapeutics using new computational tools, without having to discover them through trial and error?

We’ve put technology to work to understand proteins in ways we never could before. 

It turns out… Machine learning algorithms can generate novel sequences for proteins that have never been seen in nature. 

By training our platform on the entire compendium of protein structures and sequences found in nature—supplemented with proprietary experimental data—we can learn the generalizable rules by which a linear amino acid sequence encodes protein structure and function. 

Using what we’ve learned, we can create entirely new proteins and modalities that expand our ability to treat disease and solve complex biological challenges. This process drastically increases the success rate of and reduces the time required for drug discovery.

We call this innovation Generative Biology.

Generating proteins unlike anything that exists today.

Our machine learning algorithms analyze hundreds of millions of known proteins, looking for statistical patterns linking amino acid sequence, structure, and function.

Using these learned statistical patterns, we generate custom protein therapeutics—from short peptides to complex antibodies, enzymes, gene therapies, and yet-to-be-described protein compositions.

Shown below: Our process—from observation to analysis to the generation of novel proteins.

Charting a new future of biological engineering and medicine development.

Continual increases in computing power, coupled with an exponential rise in the production of high-throughput biological data, offers scientists a new era of drug discovery and development.

DateEvent
Machine Learning
1957Description of perceptron
1986Backpropagation paper
2000sRise of cloud computing
2010sRise of modern deep learning
2012AlexNet wins ImageNet
2014First VAE and GAN
2017Flagship explores machine learning for proteins
Medicines
1982Recombinant human insulin
1980sTransgenic mouse
1990sHTP screening
1997First humanized antibody
2000Yeast display for discovery
2018Sales of biologics surpass small molecules
Biological Engineering
1980sRise of ‘omics era
1990sInvention of NGS DNA sequencing
1997First computationally designed protein
2003De novo design of a protein fold
2010sMass DNA synthesis
2014PDB reaches 100K structures
2018Nobel prize for directed evolution