A germline knowledge based computational approach for determining antibody complementarity determining regions.

Area of Science:

• Immunology and Bioinformatics
• Computational Biology and Antibody Engineering

Background:

• Antibody framework regions (FRs) and complementarity determining regions (CDRs) are crucial for biological function and engineering.
• Existing computational methods lack automated and coherent approaches for delimiting FRs and CDRs in antibody sequences.

Purpose of the Study:

• To develop a novel computational algorithm for the automatic determination of CDRs and FRs in antibody sequences.
• To provide a robust tool for antibody sequence annotation, quality control, and engineering applications.

Main Methods:

• Developed an algorithm based on mapping relationships between mature antibody sequences and germline gene segments.
• Created a knowledge base of pre-delimited germline genes (V and J segments) to identify FRs and CDRs.
• Tested the algorithm on approximately 25,000 human antibody sequences from NCBI for robustness and accuracy.

Main Results:

• The algorithm successfully delimited over 99.7% of antibody sequences computationally.
• Only 0.28% of delimited sequences required manual checking due to somatic insertions/deletions in FRs.
• The algorithm is CDR definition independent and can be extended to various definitions (Kabat, Chothia, IMGT).

Conclusions:

• The developed algorithm offers a robust and automated solution for delimiting antibody FRs and CDRs.
• This tool enhances antibody sequence annotation, quality control, and facilitates antibody engineering.
• The algorithm's independence from specific CDR definitions and potential for integration into other applications highlight its versatility.