IgPhyML: New model and software for BCR phylogenetics

Kenneth_Hoehn · March 22, 2017, 3:57pm

IgPhyML – our program for phylogenetic analysis of antibody lineages – has been officially released. It implements a new codon substitution model that accounts for unique features of affinity maturation such as motif-sensitive hot- and cold-spot mutations, nonreversible substitution rates, and descent from a known germline sequence. It also accounts for selective effects by estimating the relative rates of nonsynonymous/synonymous mutations, which can be partitioned to estimate separate effects for FWR and CDR regions.

In the lineages we’ve tested it on, it’s shown a dramatically better fit than the standard codon model for this sort of thing (GY94), even when only adding additional mutability to symmetric WRC/GYW motifs. However, more complex motif models can be specified to include WRC/GYW, WA/TW, and/or SYC/GRS motifs, which can each be modeled as symmetric or asymmetric. It should be usable - and likely preferable to most standard phylogenetic tools - in pretty much any application of phylogenetic tree building to antibody/BCR sequences.

Let me know if you have any questions or need any help using it!

You can download IgPhyML here: https://github.com/kbhoehn/IgPhyML

and find more detail on the model in the manuscript here:

ncbi.nlm.nih.gov

A Phylogenetic Codon Substitution Model for Antibody Lineages.

KB Hoehn, G Lunter and OG Pybus, Genetics, 2017 05

Phylogenetic methods have shown promise in understanding the development of broadly neutralizing antibody lineages (bNAbs). However, the mutational process that generates these lineages, somatic hypermutation, is biased by hotspot motifs which violates important assumptions in most phylogenetic substitution models. Here, we develop a modified GY94-type substitution model that partially accounts for this context dependency while preserving independence of sites during calculation. This model shows a substantially better fit to three well-characterized bNAb lineages than the standard GY94 model. We also demonstrate how our model can be used to test hypotheses concerning the roles of different hotspot and coldspot motifs in the evolution of B-cell lineages. Further, we explore the consequences of the idea that the number of hotspot motifs, and perhaps the mutation rate in general, is expected to decay over time in individual bNAb lineages.