Institute for Protein Design

What the institute is

The Institute for Protein Design (IPD) is a research institute at the University of Washington in Seattle. It was founded on 13 April 2012, and its stated mission is to create new proteins that address challenges in medicine, technology and sustainability.^[1] David Baker, who founded the institute, is its director.^[1]

Protein design runs in the opposite direction from structure prediction. A natural protein is a chain of amino acids that folds into a particular three-dimensional shape, and that shape largely decides what the protein can do — which molecules it grips, which reactions it speeds up. Designing a protein means choosing an amino acid sequence that will fold into a shape of your own choosing and do a job that no existing protein does. The institute's premise is that this has become as much a computational problem as a biological one, which is why its work sits alongside machine learning and software development.

Who works there

The institute describes itself as bringing together biologists, computer scientists and other researchers across disciplines.^[1] The structured data on its site lists 250 to 300 employees, a figure that covers faculty, staff, translational investigators and trainees.^[1] Its core research labs are led by Baker, Bhardwaj, DiMaio and King.^[1]

Research areas

The institute lists deep learning, enzymes, peptides, minibinders, molecular devices, drug delivery, nanomaterials and membrane proteins among its research areas.^[1] The list mixes methods with subject matter. Deep learning is the computational toolkit; enzymes are proteins that catalyse chemical reactions; minibinders are small designed proteins meant to attach tightly to a chosen target; membrane proteins sit in the membranes of cells, where they are hard to study but central to signalling and transport. Its site organises the work into four areas: research, translation, software and responsible AI.^[1]

From lab to company

The institute says it has launched more than ten spinout companies that together have raised over $2 billion.^[1] It also says that, as a non-commercial entity, it receives minimal or no direct benefit from those commercial activities.^[1] The spinouts are the mechanism by which a designed protein that works in a lab can be turned into a medicine or a material that reaches users, while the academic institute keeps to research and keeps its commercial entanglements limited.

Structure prediction and RoseTTAFold

One of the institute's best-known outputs is software rather than a protein. RoseTTAFold, a three-track structure prediction network, was developed in a lab at the institute.^[2] Structure prediction takes a protein's amino acid sequence and works out the three-dimensional shape it will fold into; that shape is what determines the protein's behaviour, and knowing it is the starting point for understanding or redesigning the molecule. A tool that predicts structures accurately and is available to any researcher changes what a lab can attempt, because it replaces years of experimental work with a computation.

What it claims

The institute credits itself with the world's first computationally designed vaccine.^[1] That claim, together with the spinouts and the prediction software, is how the institute presents its case: that designing proteins from scratch is no longer a thought experiment but a route to products — vaccines, binders, enzymes and materials — built on computation first and tested in the lab second.