Efficient Inference of Clonal Families from Antibody Repertoire Sequencing Data

Core Concepts

A fast and accurate method, HILARy, is introduced to efficiently partition B-cell receptor repertoire sequencing data into clonal families, which are key to understanding the evolution and dynamics of the adaptive immune response.

Abstract

The article presents a new method, HILARy, for efficiently and accurately partitioning B-cell receptor (BCR) repertoire sequencing data into clonal families. Clonal families are lineages of related B cells stemming from the same V(D)J recombination event and are crucial for understanding the function, evolution, and dynamics of the adaptive immune response. The key insights and highlights are: The authors identify the main factors that influence the difficulty of clonal family inference: low clonality levels (prevalence) and short recombination junctions (CDR3 lengths). They develop a CDR3-based clustering method that uses a probabilistic model of recombination and selection to set an adaptive threshold, achieving high precision across different prevalence and CDR3 length regimes. To further improve performance, especially for short CDR3s and low prevalence, the authors incorporate the phylogenetic signal of shared mutations outside the CDR3 region. Benchmarking on synthetic data shows the mutations-based method outperforms state-of-the-art approaches, achieving consistently high sensitivity and precision. Applying the method to a healthy donor's repertoire, the authors find universal statistics of clonal family sizes, site frequency spectra, and selection pressures across different CDR3 length regimes, suggesting the dynamics of affinity maturation and memory formation are independent of CDR3 specificity. The proposed framework provides a robust tool for reliable partitioning of BCR repertoire data, enabling deeper insights into the adaptive immune response.

Stats

The total number of pairwise comparisons in the largest VJl classes is ~ 1010. The distribution of CDR3 lengths l shows that in-frame (productive) sequences dominate the repertoire. The prevalence (fraction of positive pairs) varies widely across VJl classes, spanning 3 orders of magnitude. The null distribution of distances between unrelated sequences, P0(x|l), becomes more peaked around 1/2 as CDR3 length l increases.

Quotes

"Clonal families are the main building blocks of the repertoire. Since members of the same family usually share their specificities, affinity maturation first competes families against each other for antigen binding in the early stages of the reaction, and then selects out the best binders within families in the later stages." "The extraordinary diversity of VDJ rearrangments can be efficiently described and quantified using probabilistic models of the recombination process as well as subsequent purifying selection." "Identifying clonal families with high accuracy is paramount in such approaches as it avoids the potential biases of different family sizes and varying levels of clonality."

Key Insights Distilled From

Combining mutation and recombination statistics to infer clonal families in antibody repertoires

by Spisak,N., D... at www.biorxiv.org 12-22-2022

https://www.biorxiv.org/content/10.1101/2022.12.22.521661v1

Deeper Inquiries

How could the proposed framework be extended to incorporate additional information, such as paired heavy and light chain data or longitudinal sampling, to further improve the accuracy of clonal family inference

To incorporate additional information like paired heavy and light chain data or longitudinal sampling into the framework for clonal family inference, several modifications and expansions can be considered: Paired Heavy and Light Chain Data: The framework can be adapted to handle paired heavy and light chain data by incorporating the unique identifiers of both chains and analyzing them together. This would provide a more comprehensive view of the antibody repertoire and allow for the identification of clonal families based on both heavy and light chain sequences. By considering the paired data, the framework can potentially improve the accuracy of clonal family inference by capturing the relationships between heavy and light chains within the same B cell clones. Longitudinal Sampling: Longitudinal sampling involves tracking changes in the antibody repertoire over time. The framework can be extended to analyze longitudinal data by incorporating time stamps or sampling intervals for each sequence. By considering the temporal aspect of the data, the framework can identify changes in clonal families over time, track the evolution of specific lineages, and detect patterns related to immune responses or disease progression. Incorporating longitudinal sampling can enhance the accuracy of clonal family inference by providing insights into the dynamics of the antibody repertoire and how clonal families evolve and diversify over time. Integration of Multi-Omics Data: To further improve accuracy, the framework could be expanded to integrate multi-omics data, such as gene expression profiles or epigenetic information related to B cell development and differentiation. By combining different types of omics data, the framework can offer a more comprehensive understanding of the factors influencing clonal family formation and evolution, leading to more accurate inference results.

What are the potential implications of the observed universal statistics of clonal family properties across CDR3 lengths for our understanding of the underlying evolutionary dynamics of the adaptive immune response

The observation of universal statistics of clonal family properties across CDR3 lengths has significant implications for our understanding of the evolutionary dynamics of the adaptive immune response: Conservation of Evolutionary Processes: The universal nature of clonal family properties suggests that the underlying evolutionary processes driving B cell repertoire diversification and selection are consistent across different CDR3 lengths. This conservation implies that the mechanisms of affinity maturation, selection, and expansion within clonal families are fundamental aspects of the adaptive immune response that remain stable regardless of the specific characteristics of the CDR3 region. Generalizability of Findings: The universal statistics indicate that findings related to clonal family structure and dynamics can be generalized across different CDR3 lengths, providing a broader applicability of research results in the field of immunology. Researchers can apply insights gained from studying specific CDR3 lengths to understand broader principles of B cell repertoire evolution and immune response dynamics. Implications for Therapeutic Development: Understanding the universal properties of clonal families can inform the development of more effective therapeutic strategies that target specific B cell lineages or leverage common evolutionary patterns to modulate immune responses. By recognizing the consistent features of clonal family dynamics, researchers can design interventions that take advantage of these universal principles to enhance immune function or treat immune-related disorders.

Could the insights gained from applying this method to healthy repertoires be leveraged to develop improved diagnostic or prognostic tools for disease-associated changes in B-cell repertoire composition and dynamics

The insights gained from applying this method to healthy repertoires can be leveraged to develop improved diagnostic or prognostic tools for disease-associated changes in B-cell repertoire composition and dynamics in the following ways: Early Disease Detection: By establishing a baseline of clonal family properties in healthy individuals, deviations from these universal statistics can be indicative of disease-associated changes in the B-cell repertoire. Monitoring alterations in clonal family structure and dynamics based on the established norms can aid in the early detection of immune-related disorders or infections. Precision Medicine: The method can be used to identify disease-specific changes in clonal families, allowing for personalized diagnostic and prognostic tools that consider individual variations in the B-cell repertoire. Tailoring treatment strategies based on the unique characteristics of clonal families can improve the efficacy of therapies and interventions for immune-related conditions. Biomarker Development: The universal statistics of clonal family properties can serve as a reference for developing biomarkers that reflect the health status of the immune system. Disease-specific alterations in clonal family composition and evolution can be used as diagnostic biomarkers to assess immune function and predict disease outcomes. By leveraging the insights from studying healthy repertoires, researchers and clinicians can enhance their ability to diagnose, monitor, and treat diseases by analyzing changes in the B-cell repertoire with a deeper understanding of the underlying evolutionary dynamics.