Antibody Somatic Hypermutation Analysis

The computational inference of clonal phylogeny from SHM patterns. Algorithms like maximum parsimony and maximum likelihood trace the evolutionary path of a B-cell clone from the unmutated common ancestor (UCA) to highly mutated progeny. This identifies intermediate ancestors and reveals the stepwise acquisition of mutations, distinguishing parallel from sequential affinity-enhancing events.

SHM is not random; it preferentially targets specific DNA motifs. WRCY (W=A/T, R=A/G, C=C, Y=C/T) and RGYW are canonical hotspots recognized by Activation-Induced Cytidine Deaminase (AID). Computational analysis quantifies mutational frequency at these motifs versus coldspots like SYC to assess whether the mutational landscape is AID-driven or shaped by selection pressure.

A fundamental metric for detecting antigen-driven selection. The R/S ratio compares the frequency of non-synonymous (replacement) mutations to synonymous (silent) mutations.

• R/S > 2.9 in CDRs: Indicates positive selection for affinity-enhancing amino acid changes.
• R/S < 2.9 in FWRs: Indicates purifying selection to preserve structural integrity. This framework, derived from the binomial probability model, distinguishes random drift from Darwinian selection.

Advanced probabilistic models like BASELINe (Bayesian Estimation of Antigen-driven Selection in Immunoglobulin Lineages) quantify selection pressure at each codon. Unlike aggregate R/S ratios, these methods estimate the posterior probability of positive or negative selection per position by comparing observed mutations to a null model of intrinsic SHM targeting, providing codon-level resolution of selective forces.

Quantifies the mutational load across the variable domain. Key metrics include:

• V-gene mutation frequency: Percentage of nucleotides mutated from germline.
• CDR3 length distribution: Longer, more mutated CDR3s often correlate with higher affinity.
• Mutational clustering: Identifying contiguous stretches of mutations that may indicate polymerase eta-mediated error-prone repair. This profile stratifies antibodies as naive, memory, or plasmablast-derived.

Identifies positions where identical amino acid substitutions occur independently across multiple clonal lineages or individuals. This convergent evolution signal strongly indicates a mutation is critical for antigen recognition. Computational methods scan large repertoire datasets to find statistically significant convergent hotspots, prioritizing them as engineering targets for affinity maturation campaigns.