Mimetic Neural Networks: A Unified Framework for Protein Design and Folding

Eliasof, Moshe; Boesen, Tue; Haber, Eldad; Keasar, Chen; Treister, Eran

doi:10.3389/fbinf.2022.715006

Cited by 12 publications

(11 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Settings: Besides JacobiConv (Wang and Zhang 2022) and EvenNet (Lei et al 2022), we also include four competitive baselines for node classification: GCNII (Chen et al 2020), TWIRLS (Yang et al 2021), EGNN (Zhou et al 2021), and PDE-GCN (Eliasof, Haber, and Treister 2021), which make full use of the topology information from different per-spectives including GNNs architecture, energy function, and PDE GNNs.…”

Section: Node Classification With Polynomial-based Methodsmentioning

confidence: 99%

PC-Conv: Unifying Homophily and Heterophily with Two-Fold Filtering

Li,

Pan,

Kang

2024

AAAI

View full text Add to dashboard Cite

Recently, many carefully designed graph representation learning methods have achieved impressive performance on either strong heterophilic or homophilic graphs, but not both. Therefore, they are incapable of generalizing well across real-world graphs with different levels of homophily. This is attributed to their neglect of homophily in heterophilic graphs, and vice versa. In this paper, we propose a two-fold filtering mechanism to mine homophily in heterophilic graphs, and vice versa. In particular, we extend the graph heat equation to perform heterophilic aggregation of global information from a long distance. The resultant filter can be exactly approximated by the Possion-Charlier (PC) polynomials. To further exploit information at multiple orders, we introduce a powerful graph convolution PC-Conv and its instantiation PCNet for the node classification task. Compared to the state-of-the-art GNNs, PCNet shows competitive performance on well-known homophilic and heterophilic graphs. Our implementation is available at https://github.com/uestclbh/PC-Conv.

show abstract

Section: Node Classification With Polynomial-based Methodsmentioning

confidence: 99%

PC-Conv: Unifying Homophily and Heterophily with Two-Fold Filtering

Li,

Pan,

Kang

2024

AAAI

View full text Add to dashboard Cite

show abstract

“…Protein folding is influenced by both local and global interactions between residues, making it a system that can be described well using a graph. We experimented with graph neural networks (GNNs) [ 9 ] and selected the one proposed in [ 30 ] as it has an interpretation of energy.…”

Section: Deriving a Surrogate Neural Networkmentioning

confidence: 99%

“…Machine Learning methods have proven to be effective tools for predicting protein structure and function, as well as designing new proteins [ 4 , 9 , 10 , 11 ], yet they avoid energy calculations all-together. Since ML methods do not use energy calculations and since they often lack interpretability and have limited explanatory power, there is no guarantee that the generated structure-sequence model optimizes some target function (e.g., stability), even approximately.…”

Section: Introductionmentioning

confidence: 99%

Protein Design Using Physics Informed Neural Networks

Omar¹,

Keasar

Ben-Sasson

et al. 2023

Biomolecules

Self Cite

View full text Add to dashboard Cite

The inverse protein folding problem, also known as protein sequence design, seeks to predict an amino acid sequence that folds into a specific structure and performs a specific function. Recent advancements in machine learning techniques have been successful in generating functional sequences, outperforming previous energy function-based methods. However, these machine learning methods are limited in their interoperability and robustness, especially when designing proteins that must function under non-ambient conditions, such as high temperature, extreme pH, or in various ionic solvents. To address this issue, we propose a new Physics-Informed Neural Networks (PINNs)-based protein sequence design approach. Our approach combines all-atom molecular dynamics simulations, a PINNs MD surrogate model, and a relaxation of binary programming to solve the protein design task while optimizing both energy and the structural stability of proteins. We demonstrate the effectiveness of our design framework in designing proteins that can function under non-ambient conditions.

show abstract

“…As we see next, choosing ϕ this way leads, in our case, to a particular form of a residual network with a double skip connection. The use of double skip connections was previously abundantly used [15,16,37] to design hyperbolic neural architectures that are motivated by the underlying neural energy of the network.…”

Section: Choice Of Nonlinearities and Potentialsmentioning

confidence: 99%

DRIP: deep regularizers for inverse problems

Eliasof,

Haber,

Treister

2023

Inverse Problems

Self Cite

View full text Add to dashboard Cite

In this paper we consider inverse problems that are mathematically ill-posed. That is, given some (noisy) data, there is more than one solution that approximately fits the data. In recent years, deep neural techniques that find the most appropriate solution, in the sense that it contains a-priori information, were developed. However, they suffer from several shortcomings. First, most techniques cannot guarantee that the solution fits the data at inference. Second, while the derivation of the techniques is inspired by the existence of a valid scalar regularization function, such techniques do not in practice rely on such a function, and therefore veer away from classical variational techniques. In this work we introduce a new family of neural regularizers for the solution of inverse problems. These regularizers are based on a variational formulation and are guaranteed to fit the data. We demonstrate their use on a number of highly ill-posed problems, from image deblurring to limited angle tomography.

show abstract

Mimetic Neural Networks: A Unified Framework for Protein Design and Folding

Cited by 12 publications

References 69 publications

PC-Conv: Unifying Homophily and Heterophily with Two-Fold Filtering

PC-Conv: Unifying Homophily and Heterophily with Two-Fold Filtering

Protein Design Using Physics Informed Neural Networks

DRIP: deep regularizers for inverse problems

Contact Info

Product

Resources

About