Deep Learning for Flexible and Site-Specific Protein Docking and Design

McPartlon, Matt; Xu, Jinbo

doi:10.1101/2023.04.01.535079

Cited by 8 publications

(6 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, GeoDock exhibited lower performance in comparison to ReplicaDock2 and AlphaFold‐Multimer (with MSAs). Like GeoDock, DockGPT (McPartlon & Xu, 2023) reported a top‐1 unbound docking success rate of 7.1% on a DB5.5 set comprising 42 targets classified as rigid, medium, or difficult. The generally lower success rates for unbound docking highlight the current challenges in docking approaches to capture backbone conformational changes.…”

Section: Resultsmentioning

confidence: 99%

“…Currently, GeoDock employs a pretrained language model to integrate sequence information. Nevertheless, ablation studies conducted by DockGPT (McPartlon & Xu, 2023) suggested that incorporating ESM‐1b (Rives et al, 2021) embeddings for each chain does not markedly enhance docking performance. This lack of improvement may be attributed to the absence of interchain relations in the pretrained language model, given that the training data consists solely of single‐chain sequences.…”

Section: Discussionmentioning

confidence: 99%

“…Inspired by the work from McPartlon and Xu (2023), we provided a small fraction of interchain contact (distance between

{\mathrm{C}}_{\beta }

atoms under 10 Å) information as a binary embedding

{C}_{ij}\in \left\{0,1\right\}

concatenated to the edge embeddings during training. Specifically, for each training example, we generated a random integer

i\in \left[0,3\right]

and randomly sampled

i

interchain contacts.…”

Section: Methodsmentioning

confidence: 99%

“…Due to the rigid‐body assumption, these methods are not capable of predicting the conformational changes upon protein binding. More recently, McPartlon and Xu (2023) proposed DockGPT for flexible and site‐specific protein docking and design, demonstrating high docking success rates when partial binding site information is provided.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Flexible protein–protein docking with a multitrack iterative transformer

Chu,

Ruffolo,

Harmalkar

et al. 2024

Protein Science

View full text Add to dashboard Cite

Conventional protein‐protein docking algorithms usually rely on heavy candidate sampling and re‐ranking, but these steps are time‐consuming and hinder applications that require high‐throughput complex structure prediction, e.g., structure‐based virtual screening. Existing deep learning methods for protein‐protein docking, despite being much faster, suffer from low docking success rates. In addition, they simplify the problem to assume no conformational changes within any protein upon binding (rigid docking). This assumption precludes applications when binding‐induced conformational changes play a role, such as allosteric inhibition or docking from uncertain unbound model structures. To address these limitations, we present GeoDock, a multi‐track iterative transformer network to predict a docked structure from separate docking partners. Unlike deep learning models for protein structure prediction that input multiple sequence alignments (MSAs), GeoDock inputs just the sequences and structures of the docking partners, which suits the tasks when the individual structures are given. GeoDock is flexible at the protein residue level, allowing the prediction of conformational changes upon binding. On the DIPS test set, GeoDock achieves a 43% top‐1 success rate, outperforming all other tested methods. However, in the standard DIPS train/test splits, we discovered contamination of close homologs in the training set. After decontaminating the training set, the success rate is 31%. On the DB5.5 test set and a benchmark dataset of antibody‐antigen complexes, GeoDock outperforms the deep learning models trained using the same dataset but falls behind most of the conventional methods and AlphaFold‐Multimer. GeoDock attains an average inference speed of under one second on a single GPU, enabling its application in large‐scale structure screening. Although binding‐induced conformational changes are still a challenge owing to limited training and evaluation data, our architecture sets up the foundation to capture this backbone flexibility. Code and a demonstration Jupyter notebook are available at https://github.com/Graylab/GeoDock.This article is protected by copyright. All rights reserved.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

“…Inspired by the work from McPartlon and Xu (2023), we provided a small fraction of interchain contact (distance between

{\mathrm{C}}_{\beta }

atoms under 10 Å) information as a binary embedding

{C}_{ij}\in \left\{0,1\right\}

concatenated to the edge embeddings during training. Specifically, for each training example, we generated a random integer

i\in \left[0,3\right]

and randomly sampled

i

interchain contacts.…”

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Flexible protein–protein docking with a multitrack iterative transformer

Chu,

Ruffolo,

Harmalkar

et al. 2024

Protein Science

View full text Add to dashboard Cite

show abstract

“…In addition, the docking success rates of these methods are lower than the conventional docking methods. More recently, McPartlon & Xu [57] proposed DockGPT for flexible and site-specific protein docking and design, demonstrating high docking success rates when partial binding site information is provided.…”

Section: Introductionmentioning

confidence: 99%

Flexible Protein-Protein Docking with a Multi-Track Iterative Transformer

Chu

Ruffolo

Harmalkar

et al. 2023

Preprint

View full text Add to dashboard Cite

Conventional protein-protein docking algorithms usually rely on heavy candidate sampling and re-ranking, but these steps are time-consuming and hinder applications that require high-throughput complex structure prediction, e.g., structure-based virtual screening. Existing deep learning methods for protein-protein docking, despite being much faster, suffer from low docking success rates. In addition, they simplify the problem to assume no conformational changes within any protein upon binding (rigid docking). This assumption precludes applications when binding-induced conformational changes play a role, such as allosteric inhibition or docking from uncertain unbound model structures. To address these limitations, we present GeoDock, a multi-track iterative transformer network to predict a docked structure from separate docking partners. Unlike deep learning models for protein structure prediction that input multiple sequence alignments (MSAs), GeoDock inputs just the sequences and structures of the docking partners, which suits the tasks when the individual structures are given. GeoDock is flexible at the protein residue level, allowing the prediction of conformational changes upon binding. For a benchmark set of rigid targets, GeoDock obtains a 41% success rate, outperforming all the other tested methods. For a more challenging benchmark set of flexible targets, GeoDock achieves a similar number of top-model successes as the traditional method ClusPro, but fewer than ReplicaDock2. GeoDock attains an average inference speed of under one second on a single GPU, enabling its application in large-scale structure screening. Although binding-induced conformational changes are still a challenge owing to limited training and evaluation data, our architecture sets up the foundation to capture this backbone flexibility. Code and a demonstration Jupyter notebook are available at https://github.com/Graylab/GeoDock.

show abstract

Applications and challenges in designing VHH-based bispecific antibodies: leveraging machine learning solutions

Mullin,

McClory,

Haynes

et al. 2024

mAbs

View full text Add to dashboard Cite

The development of bispecific antibodies that bind at least two different targets relies on bringing together multiple binding domains with different binding properties and biophysical characteristics to produce a drug-like therapeutic. These building blocks play an important role in the overall quality of the molecule and can influence many important aspects from potency and specificity to stability and half-life. Single-domain antibodies, particularly camelid-derived variable heavy domain of heavy chain (VHH) antibodies, are becoming an increasingly popular choice for bispecific construction due to their single-domain modularity, favorable biophysical properties, and potential to work in multiple antibody formats. Here, we review the use of VHH domains as building blocks in the construction of multispecific antibodies and the challenges in creating optimized molecules. In addition to exploring traditional approaches to VHH development, we review the integration of machine learning techniques at various stages of the process. Specifically, the utilization of machine learning for structural prediction, lead identification, lead optimization, and humanization of VHH antibodies.

show abstract

Deep Learning for Flexible and Site-Specific Protein Docking and Design

Cited by 8 publications

References 42 publications

Flexible protein–protein docking with a multitrack iterative transformer

Flexible protein–protein docking with a multitrack iterative transformer

Flexible Protein-Protein Docking with a Multi-Track Iterative Transformer

Applications and challenges in designing VHH-based bispecific antibodies: leveraging machine learning solutions

Contact Info

Product

Resources

About