ParaSurf: A Surface-Based Deep Learning Approach for Paratope-Antigen Interaction Prediction

Salvato in:
Dettagli Bibliografici
Pubblicato in:bioRxiv (Jan 30, 2025)
Autore principale: Papadopoulos, Angelos Michael
Altri autori: Axenopoulos, Apostolos, Iatrou, Anastasia, Stamatopoulos, Kostas, Alvarez, Federico, Daras, Petros
Pubblicazione:
Cold Spring Harbor Laboratory Press
Soggetti:
Accesso online:Citation/Abstract
Full Text - PDF
Full text outside of ProQuest
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!

MARC

LEADER 00000nab a2200000uu 4500
001 3161603203
003 UK-CbPIL
022 |a 2692-8205 
024 7 |a 10.1101/2024.12.16.628621  |2 doi 
035 |a 3161603203 
045 0 |b d20250130 
100 1 |a Papadopoulos, Angelos Michael 
245 1 |a ParaSurf: A Surface-Based Deep Learning Approach for Paratope-Antigen Interaction Prediction 
260 |b Cold Spring Harbor Laboratory Press  |c Jan 30, 2025 
513 |a Working Paper 
520 3 |a Motivation: Identifying antibody binding sites, is crucial for developing vaccines and therapeutic antibodies, processes that are time-consuming and costly. Accurate prediction of the paratope's binding site can speed up the development by improving our understanding of antibody-antigen interactions. Results: We present ParaSurf, a deep learning model that significantly enhances paratope prediction by incorporating both surface geometric and non-geometric factors. Trained and tested on three prominent antibody-antigen benchmarks, ParaSurf achieves state-of-the-art results across nearly all metrics. Unlike models restricted to the variable region, ParaSurf demonstrates the ability to accurately predict binding scores across the entire Fab region of the antibody. Additionally, we conducted an extensive analysis using the largest of the three datasets employed, focusing on three key components: (1) a detailed evaluation of paratope prediction for each Complementarity-Determining Region loop, (2) the performance of models trained exclusively on the heavy chain, and (3) the results of training models solely on the light chain without incorporating data from the heavy chain. Availability and Implementation: Source code for ParaSurf, along with the datasets used, preprocessing pipeline, and trained model weights, are freely available at https://github.com/aggelos-michael-papadopoulos/ParaSurf. Contact: [email protected], [email protected]Supplementary information: Supplementary data are provided as a separate file with this submission.Competing Interest StatementThe authors have declared no competing interest.Footnotes* We have created a new dataset, which is actually the pool of the 3 benchmark datasets; PECAN + Paragraph Expanded + MIPE to showcase our model's best performance (changes also shown in Supplementary material). Also some minor corrections on the text took place* https://github.com/aggelos-michael-papadopoulos/ParaSurf 
653 |a Variable region 
653 |a Datasets 
653 |a Deep learning 
653 |a Antigens 
653 |a Models 
653 |a Antigen-antibody interactions 
653 |a Antibodies 
653 |a Complementarity-determining region 
653 |a Predictions 
700 1 |a Axenopoulos, Apostolos 
700 1 |a Iatrou, Anastasia 
700 1 |a Stamatopoulos, Kostas 
700 1 |a Alvarez, Federico 
700 1 |a Daras, Petros 
773 0 |t bioRxiv  |g (Jan 30, 2025) 
786 0 |d ProQuest  |t Biological Science Database 
856 4 1 |3 Citation/Abstract  |u https://www.proquest.com/docview/3161603203/abstract/embedded/L8HZQI7Z43R0LA5T?source=fedsrch 
856 4 0 |3 Full Text - PDF  |u https://www.proquest.com/docview/3161603203/fulltextPDF/embedded/L8HZQI7Z43R0LA5T?source=fedsrch 
856 4 0 |3 Full text outside of ProQuest  |u https://www.biorxiv.org/content/10.1101/2024.12.16.628621v2