24BBR_Bhive-plus

This application proposes to build BHive+, an integrated software ecosystem dedicated for systems immunology of antibodies and B-cells. Antibodies, produced by specialised immune cells known as B-cells, play a crucial role in our immune response by neutralising pathogens and recruiting effector cells to clear these pathogens, thereby protecting us from infections. They can also act as receptors on the B-cell surface to detect target molecules (“antigens”). We have previously generated a user-friendly webserver (BRepertoire) for statistical analyses of large-scale antibody sequence “repertoires” to understand how diverse antibody responses are achieved across individuals facing different immune challenges. This has further been complemented by a suite of computational tools (sciCSR, BrepPhylo, BrepConvert) we developed for detailed analyses of how these antibody repertoires were developed in vivo, as well as the dynamics of B-cell maturation. Whilst these tools, together forming the collection “BHive”, tackle specific aspects of B-cell and antibody-related data analysis, the immunology community has unmet data analysis needs on the annotation of immunoglobulin sequences, structures, and B-cell subtypes. The community also lacks easy-to-use software tools to extract novel insights from high-dimensional data, and apply cutting-edge artificial intelligence and machine learning methods to devise B-cell therapies and the design of antibody therpaeutics. This is particularly timely as the COVID-19 pandemic and recent novel approaches in cancer immunotherapy demonstrate that therapeutic antibodies are a vital pharmaceutical resource, with thousands in development, primarily IgG1, but none IgA or IgE. Understanding the functions of different classes of antibodies could expand their potential utility, such as IgE for cancer therapies or IgA for treating mucosal disorders in the airways and the gut. In this programme, we aim to leverage our expertise in bioinformatics and immunology to address gaps in the analysis of single-cell transcriptomics and immunophenotyping of B-cells, as well as understanding of the structure-function relationship of antibodies. Addressing the unmet needs of different user communities, we plan to create: A specialized, unified data repository (BrepData) clarifying annotations of antibodies and B-cells, facilitating experimental and computational immunologists. A new package large-scale B-cell cytometry data analysis (BrepPheno) to facilitate cellular immunologists to characterise B-cell subsets. Software for structural biologists to model full-length antibody structures and their energetic analyses (BrepPoses). Software to democratisie the application of state-of-the-art protein language models in antibody engineering and specificity prediction (BrepML) to antibody discovery scientists. An integrated software ecosystem, BHive+, featuring these new implementations and the existing BHive tools for all the aforementioned user communities. All proposed additions enable new data analyses in B-cell immunology and antibody science which are challenging without significant domain knowledge and expertise. Our software will be provided as both graphical user interfaces and standalone R packages. User support will be provided via documentation and code notebooks oriented to immunologists, face-to-face workshops and an online community promoting user inputs in software updates and the creation of new use-cases focusing on community need. We foresee, through this programme, BHive+ will become a central hub of computational resources dedicated to B-cell immunology, enabling multimodal data analysis for B-cell and antibody-related research, and establishing itself as a “one-stop shop” for B-cell systems immunology.