A Sequence-Based Antibody Paratope Prediction Model Through Combing Local-Global Information and Partner Features

Lü, Shan; Li, Yuguang; Nan, Xiaofei; Zhang, Shoutao

doi:10.1007/978-3-030-91415-8_16

Cited by 2 publications

(1 citation statement)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These algorithms predict those residues most likely to be directly involved in antigen binding from the antibody sequence. Many such predictors are available and their accuracy has been steadily increasing (69)(70)(71)(72)(73), and one, called Parapred (69), can be accessed directly on our webserver. Whether predicted or experimentally determined, functionally relevant positions can be excluded from the design, as we have done for Nb.b201 in this work.…”

Section: Discussionmentioning

confidence: 99%

Automated optimisation of solubility and conformational stability of antibodies and proteins

Rosace

Bennett

Oeller³

et al. 2022

Preprint

View full text Add to dashboard Cite

Developability issues, such as aggregation, low expression yield, or instability over long-term storage often hamper the development of biologics. Developability is underpinned by a complex interplay of biophysical properties, including conformational stability and solubility. Advances have been made in the optimisation of individual properties, however, multi-trait optimisation remains highly challenging, as mutations that improve one property often negatively impact the others. Here, we introduce a fully automated computational strategy for the rational design of conformationally stable and soluble protein and antibody variants harbouring multiple mutations. This pipeline, called CamSol Combination, leverages phylogenetic information to reduce false positive predictions, and combines a rapid method of predicting solubility changes with an empirical energy function. We experimentally validate the method’s predictions on a nanobody isolated with yeast-display from a synthetic library, by producing 12 designs ranging from single-point mutants to a quadruple mutant. All 12 designed variants retained antigen binding, had improved thermal stability, and decreased propensity to precipitate (increased relative solubility), and 8 also had reduced cross-reactivity. The melting temperature was improved by 13.6 ºC with 4 rationally designed mutations. We make the method available as a webserver at www-cohsoftware.ch.cam.ac.ukSignificanceProtein-based biologics, such as antibodies and enzymes, are crucial reagents in research, industrial biotechnology, and diagnostics, and are increasingly used to treat a wide range of diseases as biotherapeutics. Often, biologics with suitable functionality are discovered, but their development into practically useful molecules is impeded by developability issues. Conformational stability and solubility are arguably the most important biophysical properties underpinning developability potential, as they determine colloidal stability and aggregation, and correlate with expression yield and poly-specificity. Here we introduce and experimentally validate a fully automated computational method and associated webserver for the rational design of proteins and antibodies with improved stability and solubility. Computational methods are rapid, inexpensive, and have no material requirements, which makes their implementation into biologic development pipelines particularly attractive.

show abstract

Section: Discussionmentioning

confidence: 99%

Automated optimisation of solubility and conformational stability of antibodies and proteins

Rosace

Bennett

Oeller³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

PMSFF: Improved Protein Binding Residues Prediction through Multi-Scale Sequence-Based Feature Fusion Strategy

Li,

Nan,

Zhang

et al. 2024

Biomolecules

View full text Add to dashboard Cite

Proteins perform different biological functions through binding with various molecules which are mediated by a few key residues and accurate prediction of such protein binding residues (PBRs) is crucial for understanding cellular processes and for designing new drugs. Many computational prediction approaches have been proposed to identify PBRs with sequence-based features. However, these approaches face two main challenges: (1) these methods only concatenate residue feature vectors with a simple sliding window strategy, and (2) it is challenging to find a uniform sliding window size suitable for learning embeddings across different types of PBRs. In this study, we propose one novel framework that could apply multiple types of PBRs Prediciton task through Multi-scale Sequence-based Feature Fusion (PMSFF) strategy. Firstly, PMSFF employs a pre-trained language model named ProtT5, to encode amino acid residues in protein sequences. Then, it generates multi-scale residue embeddings by applying multi-size windows to capture effective neighboring residues and multi-size kernels to learn information across different scales. Additionally, the proposed model treats protein sequences as sentences, employing a bidirectional GRU to learn global context. We also collect benchmark datasets encompassing various PBRs types and evaluate our PMSFF approach to these datasets. Compared with state-of-the-art methods, PMSFF demonstrates superior performance on most PBRs prediction tasks.

show abstract

A Sequence-Based Antibody Paratope Prediction Model Through Combing Local-Global Information and Partner Features

Cited by 2 publications

References 39 publications

Automated optimisation of solubility and conformational stability of antibodies and proteins

Automated optimisation of solubility and conformational stability of antibodies and proteins

PMSFF: Improved Protein Binding Residues Prediction through Multi-Scale Sequence-Based Feature Fusion Strategy

Contact Info

Product

Resources

About