Related Experiment Video
Updated: May 13, 2025
Identification of Mouse and Human Antibody Repertoires by Next-Generation Sequencing
Published on: March 15, 2019
AbSet: A Standardized Data Set of Antibody Structures for Machine Learning Applications
Diego S Almeida1,2, Matheus V Almeida1, Jean V Sampaio1,2
1Laboratory of Structural and Functional Biology Applied to Biopharmaceuticals, Fundação Oswaldo Cruz, Fiocruz Ceará, Eusébio 61773-270, Brazil.
A new dataset, AbSet, offers over 800,000 antibody structures and molecular descriptors to improve machine learning models for therapeutic antibody development. This resource addresses limitations in existing structural data for antibody-antigen complexes.
Area of Science:
- Biochemistry
- Structural Biology
- Computational Biology
Background:
- Machine learning (ML) is crucial for therapeutic antibody development, relying on sequence and structural data.
- Existing structural datasets are limited, particularly for antibody-antigen complexes, and often lack standardization and essential molecular descriptors.
- This data scarcity hinders the development of robust ML models for antibody design and engineering.
Purpose of the Study:
- To introduce AbSet, a comprehensive and curated dataset of antibody structures and molecular descriptors.
- To address the limitations of existing structural data for machine learning applications in therapeutic antibody development.
- To provide a valuable resource for researchers aiming to improve antibody design and predict antibody-antigen interactions.
Main Methods:
- Systematic retrieval of antibody structures from the Protein Data Bank (PDB).
- Application of rigorous standardization protocols to ensure data consistency.
- Expansion of the dataset through large-scale protein-protein docking to generate in silico antibody-antigen complexes.
- Quality classification of generated models based on structural similarity to experimental data.
Main Results:
- Creation of AbSet, a dataset containing over 800,000 antibody structures and molecular descriptors.
- Inclusion of both experimentally determined and computationally generated antibody-antigen complexes.
- Development of a quality assessment system for structural models, enabling the creation of decoy sets and high-confidence data.
- Public availability of the AbSet dataset and associated scripts.
Conclusions:
- AbSet provides a standardized, high-quality, and expanded resource for machine learning in therapeutic antibody development.
- The dataset facilitates the creation of improved ML models by offering diverse structural data and molecular descriptors.
- AbSet supports the advancement of antibody engineering and the prediction of antibody-antigen interactions.
More Related Videos
04:24Author Spotlight: Advancing Biotherapeutic Mass Calculation by Introducing mAbScale, a Python-Based Desktop Application
Published on: June 16, 2023
00:09Single-cell Screening Method for the Selection and Recovery of Antibodies with Desired Specificities from Enriched Human Memory B Cell Populations
Published on: August 22, 2019
Related Concept Videos
01:10Antibody Structure
Antibodies, also known as immunoglobulins (Ig), are essential players of the adaptive immune system. These antigen-binding proteins are produced by B cells and make up 20 percent of the total blood plasma by weight. In mammals, antibodies fall into five different classes, which each elicits a different biological response upon antigen binding.
The Y-Shaped Structure of Antibodies Consists of Four Polypeptide Chains
Antibodies consist of four polypeptide chains: two identical heavy...
01:25Antibody Structure and Classes
The basic structure of an antibody consists of four protein chains: two identical heavy chains and two identical light chains. These chains are held together by disulfide bonds and other non-covalent interactions, forming a Y-shaped structure.