De novo protein design by inversion of the AlphaFold structure prediction network

Abstract

De novo protein design enhances our understanding of the principles that govern protein folding and interactions, and has the potential to revolutionize biotechnology through the engineering of novel protein functionalities. Despite recent progress in computational design strategies, de novo design of protein structures remains challenging, given the vast size of the sequence-structure space. AlphaFold2 (AF2), a state-of-the-art neural network architecture, achieved remarkable accuracy in predicting protein structures from amino acid sequences. This raises the question whether AF2 has learned the principles of protein folding sufficiently for de novo design. Here, we sought to answer this question by inverting the AF2 network, using the prediction weight set and a loss function to bias the generated sequences to adopt a target fold. Initial design trials resulted in de novo designs with an overrepresentation of hydrophobic residues on the protein surface compared to their natural protein family, requiring additional surface optimization. In silico validation of the designs showed protein structures with the correct fold, a hydrophilic surface and a densely packed hydrophobic core. In vitro validation showed that several designs were folded and stable in solution with high melting temperatures. In summary, our design workflow solely based on AF2 does not seem to fully capture basic principles of de novo protein design, as observed in the protein surface's hydrophobic vs. hydrophilic patterning. However, with minimal post-design intervention, these pipelines generated viable sequences as assessed experimental characterization. Thus such pipelines show the potential to contribute to solving outstanding challenges in de novo protein design.

Cyclic peptide structure prediction and design using AlphaFold

“…We started with macrocycles composed of 7–10 amino acids and enumerated 48,000 hallucinated models for each size. We clustered the resulting structures from these large sampling runs using torsion bin-based clustering described earlier, and identified 9941, 13405, 19705, and 22206 unique structural clusters for 7-mers, 8-mers, 9-mers, and 10-mer cyclic peptides, respectively (Figure 3A).…”

“…Many of these hallucinated sequences demonstrated promising folding propensity in these calculations, with 114 7-mers, 186 8-mers, 139 9-mers, and 76 10-mer sequences showing Rosetta P near values greater than 0.6. We selected one hallucinated design model per size between 7–10 amino acids with AlphaFold pLDDT > 0.9 and Rosetta P near > 0.9 for further experimental validation and structural characterization. All four selected designs lack regular secondary structures but are stabilized by extensive intramolecular backbone-to-backbone and backbone-to-sidechain hydrogen bonding.…”

Efficient and scalablede novoprotein design using a relaxed sequence space

“…The copyright holder for this this version posted February 25, 2023. ; https://doi.org/10.1101/2023.02.24.529906 doi: bioRxiv preprint in the sequence update loop 5,12,14 . Though this approach worked well for models predicting distribution of distances for every pair of positions, such as TrRosetta, the approach did not work well in more recent models such as AlphaFold, where single amino acid changes could radically alter the predicted structure, resulting in unstable and inefficient optimization.…”

“…This approach may also fail to identify a solution, similar to issues faced by numerical solutions to mathematical problems in a random Monte Carlo fashion. To improve the likelihood of generating accurate predictions, researchers have proposed using sequence gradients obtained by inverting structure prediction networks 12 . However, when updating the gradients, a major issue arises due to the discrete, one-hot encoded representation of sequences.…”

“…: 1, A) represents each amino acid or nucleotide as a binary vector with all zeros, except for a single ‘1’, indicating the position of the corresponding residue. While the obtained gradient from the backpropagation through the structure prediction network is non discrete, previous approaches tried to regain discreteness by a ‘straight-through estimator’ 13 in the sequence update loop 5,12,14 . Though this approach worked well for models predicting distribution of distances for every pair of positions, such as TrRosetta, the approach did not work well in more recent models such as AlphaFold, where single amino acid changes could radically alter the predicted structure, resulting in unstable and inefficient optimization.…”