conf/P2PXML_Structure/README.md

# P2PXML-PDB Dataset

To support the development of antibody-antigen binding affinity prediction frameworks, we curated the largest and most generalized datasets for antibody-antigen binding affinity prediction in the literature. To this end, P2PXML-PDB is the curated protein structure dataset. 

## Abstract from the accompanied paper [https://www.biorxiv.org/content/early/2024/06/10/2024.06.09.598103](Paper):

In the context of drug development, the efficacy of drugs designed using biologics depends on the strength of the interactions between the biologics and the target molecule. In practise, these strengths of the interactions are measured using binding affinity and thus, understanding the binding affinity between an engineered antibody and its target antigen is a critical aspect of antibody engineering as it reflects the overall effectiveness of the antibody in inhibiting the activity of the target antigen. In theory, potential candidate antibodies have a higher binding affinity towards the target antigen. Currently, techniques such as Molecular docking and Molecular dynamics are utilized in quantifying the binding affinity. However, owing to the computational complexity of the aforementioned techniques, running simulations for large antibodies/antigens remains a daunting task. Despite the commendable improvements in deep learning-based binding affinity prediction, such approaches are highly dependent on the quality of the antibody-antigen structures and they tend to overlook the importance of capturing the evolutionary details of proteins upon mutation. Further, most of the existing datasets for the task only include antibody-antigen pairs related to one antigen variant and thus, not suitable for developing comprehensive data-driven approaches. To circumvent the said complexities, we first curate the largest and most generalized datasets for antibody-antigen binding affinity prediction, consisting both protein sequences and structures, in the literature and then propose a deep geometric neural network comprising a structure-based model and a sequence-based model that considers both atomistic and evolutionary details when predicting the binding affinity. The proposed framework exhibited a $10\%$ improvement in mean absolute error compared to the state-of-the-art models while showing a very strong correlation between the predictions and target values. We release the datasets and code publicly [https://drug-discovery-entc.github.io/p2pxml/](web-page) to support the development of antibody-antigen binding affinity prediction frameworks for the benefit of science and society.

## Dataset Overview

  Data include binding affinity values (in terms of IC50 value) between a diverse sets of antibodies and antigens. To our knowledge, this dataset is the largest, publicly-available dataset that contains antibody structures, antigen structures and quantitative measurements and provides an opportunity to serve as a benchmark to evaluate generalized and data-driven antibody-antigen binding affinity prediction frameworks.

P2PXML-PDB dataset has two folders, named "antibodies" and "antigens", and a csv file with the following columns entries:

| **Variable Name**        | **Description**    |
| ------------------------ | ------------- | 
| Ab     | Antibody name | 
| Ag      | Antigen name      |  
| IC50| The binding affinity between the corresponding antibody and antigen  | 
| log(IC50)     | Logarithmic (to base 10) value of the corresponding IC50 | 

The antibody and antigen structures (in PDB format), which are corresponding to the names in the "Ab" and "Ag" columns, are in the respective folders.

## Citation

@article {Bandara2024.06.09.598103,
	author = {Bandara, Nuwan Sriyantha and Premathilaka, Dasun and Chandanayake, Sachini and Hettiarachchi, Sahan and Varenthirarajah, Vithurshan and Munasinghe, Aravinda and Madhawa, Kaushalya and Charles, Subodha},
	title = {Deep Geometric Framework to Predict Antibody-Antigen Binding Affinity},
	elocation-id = {2024.06.09.598103},
	year = {2024},
	doi = {10.1101/2024.06.09.598103},
	publisher = {Cold Spring Harbor Laboratory},
	URL = {https://www.biorxiv.org/content/early/2024/06/10/2024.06.09.598103},
	eprint = {https://www.biorxiv.org/content/early/2024/06/10/2024.06.09.598103.full.pdf},
	journal = {bioRxiv}
}


## License


Shield: [![CC BY-NC-SA 4.0][cc-by-nc-sa-shield]][cc-by-nc-sa]

This work is licensed under a
[Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License][cc-by-nc-sa].

[![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]

[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/
[cc-by-nc-sa-image]: https://licensebuttons.net/l/by-nc-sa/4.0/88x31.png
[cc-by-nc-sa-shield]: https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg


## Disclaimer

DISTRIBUTION STATEMENT A. Approved for public release. Distribution is unlimited.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the affiliated institutions of the author(s). 

© 2024 P2PXML Team.