De novo design of immunoglobulin-like domains.

Antibodies, and antibody derivatives such as nanobodies, contain immunoglobulin-like (Ig) ß-sandwich scaffolds which anchor the hypervariable antigen-binding loops and constitute the largest growing class of drugs. Current engineering strategies for this class of compounds rely on naturally existing Ig frameworks, which can be hard to modify and have limitations in manufacturability, designability and range of action. Here, we develop design rules for the central feature of the Ig fold architecture-the non-local cross-ß structure connecting the two ß-sheets-and use these to design highly stable Ig domains de novo, confirm their structures through X-ray crystallography, and show they can correctly scaffold functional loops. Our approach opens the door to the design of antibody-like scaffolds with tailored structures and superior biophysical properties.

Large-scale design and refinement of stable proteins using sequence-only models.

Engineered proteins generally must possess a stable structure in order to achieve their designed function. Stable designs, however, are astronomically rare within the space of all possible amino acid sequences. As a consequence, many designs must be tested computationally and experimentally in order to find stable ones, which is expensive in terms of time and resources. Here we use a high-throughput, low-fidelity assay to experimentally evaluate the stability of approximately 200,000 novel proteins. These include a wide range of sequence perturbations, providing a baseline for future work in the field. We build a neural network model that predicts protein stability given only sequences of amino acids, and compare its performance to the assayed values. We also report another network model that is able to generate the amino acid sequences of novel stable proteins given requested secondary sequences. Finally, we show that the predictive model-despite weaknesses including a noisy data set-can be used to substantially increase the stability of both expert-designed and model-generated proteins.

De novo protein design by deep network hallucination.

There has been considerable recent progress in protein structure prediction using deep neural networks to predict inter-residue distances from amino acid sequences1-3. Here we investigate whether the information captured by such networks is sufficiently rich to generate new folded proteins with sequences unrelated to those of the naturally occurring proteins used in training the models. We generate random amino acid sequences, and input them into the trRosetta structure prediction network to predict starting residue-residue distance maps, which, as expected, are quite featureless. We then carry out Monte Carlo sampling in amino acid sequence space, optimizing the contrast (Kullback-Leibler divergence) between the inter-residue distance distributions predicted by the network and background distributions averaged over all proteins. Optimization from different random starting points resulted in novel proteins spanning a wide range of sequences and predicted structures. We obtained synthetic genes encoding 129 of the network-'hallucinated' sequences, and expressed and purified the proteins in Escherichia coli; 27 of the proteins yielded monodisperse species with circular dichroism spectra consistent with the hallucinated structures. We determined the three-dimensional structures of three of the hallucinated proteins, two by X-ray crystallography and one by NMR, and these closely matched the hallucinated models. Thus, deep networks trained to predict native protein structures from their sequences can be inverted to design new proteins, and such networks and methods should contribute alongside traditional physics-based models to the de novo design of proteins with new functions.

De novo design of a non-local ß-sheet protein with high stability and accuracy.

ß-sheet proteins carry out critical functions in biology, and hence are attractive scaffolds for computational protein design. Despite this potential, de novo design of all-ß-sheet proteins from first principles lags far behind the design of all-α or mixed-αß domains owing to their non-local nature and the tendency of exposed ß-strand edges to aggregate. Through study of loops connecting unpaired ß-strands (ß-arches), we have identified a series of structural relationships between loop geometry, side chain directionality and ß-strand length that arise from hydrogen bonding and packing constraints on regular ß-sheet structures. We use these rules to de novo design jellyroll structures with double-stranded ß-helices formed by eight antiparallel ß-strands. The nuclear magnetic resonance structure of a hyperthermostable design closely matched the computational model, demonstrating accurate control over the ß-sheet structure and loop geometry. Our results open the door to the design of a broad range of non-local ß-sheet protein structures.

Global analysis of protein folding using massively parallel design, synthesis, and testing.

Proteins fold into unique native structures stabilized by thousands of weak interactions that collectively overcome the entropic cost of folding. Although these forces are "encoded" in the thousands of known protein structures, "decoding" them is challenging because of the complexity of natural proteins that have evolved for function, not stability. We combined computational protein design, next-generation gene synthesis, and a high-throughput protease susceptibility assay to measure folding and stability for more than 15,000 de novo designed miniproteins, 1000 natural proteins, 10,000 point mutants, and 30,000 negative control sequences. This analysis identified more than 2500 stable designed proteins in four basic folds-a number sufficient to enable us to systematically examine how sequence determines folding and stability in uncharted protein space. Iteration between design and experiment increased the design success rate from 6% to 47%, produced stable proteins unlike those found in nature for topologies where design was initially unsuccessful, and revealed subtle contributions to stability as designs became increasingly optimized. Our approach achieves the long-standing goal of a tight feedback cycle between computation and experiment and has the potential to transform computational protein design into a data-driven science.

Principles for designing proteins with cavities formed by curved ß sheets.

Active sites and ligand-binding cavities in native proteins are often formed by curved ß sheets, and the ability to control ß-sheet curvature would allow design of binding proteins with cavities customized to specific ligands. Toward this end, we investigated the mechanisms controlling ß-sheet curvature by studying the geometry of ß sheets in naturally occurring protein structures and folding simulations. The principles emerging from this analysis were used to design, de novo, a series of proteins with curved ß sheets topped with α helices. Nuclear magnetic resonance and crystal structures of the designs closely match the computational models, showing that ß-sheet curvature can be controlled with atomic-level accuracy. Our approach enables the design of proteins with cavities and provides a route to custom design ligand-binding and catalytic sites.