Representing and extending ensembles of parsimonious evolutionary histories with a directed acyclic graph.

In many situations, it would be useful to know not just the best phylogenetic tree for a given data set, but the collection of high-quality trees. This goal is typically addressed using Bayesian techniques, however, current Bayesian methods do not scale to large data sets. Furthermore, for large data sets with relatively low signal one cannot even store every good tree individually, especially when the trees are required to be bifurcating. In this paper, we develop a novel object called the "history subpartition directed acyclic graph" (or "history sDAG" for short) that compactly represents an ensemble of trees with labels (e.g. ancestral sequences) mapped onto the internal nodes. The history sDAG can be built efficiently and can also be efficiently trimmed to only represent maximally parsimonious trees. We show that the history sDAG allows us to find many additional equally parsimonious trees, extending combinatorially beyond the ensemble used to construct it. We argue that this object could be useful as the "skeleton" of a more complete uncertainty quantification.

Many-core algorithms for high-dimensional gradients on phylogenetic trees.

The rapid growth in genomic pathogen data spurs the need for efficient inference techniques, such as Hamiltonian Monte Carlo (HMC) in a Bayesian framework, to estimate parameters of these phylogenetic models where the dimensions of the parameters increase with the number of sequences $N$. HMC requires repeated calculation of the gradient of the data log-likelihood with respect to (wrt) all branch-length-specific (BLS) parameters that traditionally takes $\mathcal{O}(N^2)$ operations using the standard pruning algorithm. A recent study proposes an approach to calculate this gradient in $\mathcal{O}(N)$, enabling researchers to take advantage of gradient-based samplers such as HMC. The CPU implementation of this approach makes the calculation of the gradient computationally tractable for nucleotide-based models but falls short in performance for larger state-space size models, such as codon models. Here, we describe novel massively parallel algorithms to calculate the gradient of the log-likelihood wrt all BLS parameters that take advantage of graphics processing units (GPUs) and result in many fold higher speedups over previous CPU implementations. We benchmark these GPU algorithms on three computing systems using three evolutionary inference examples: carnivores, dengue and yeast, and observe a greater than 128-fold speedup over the CPU implementation for codon-based models and greater than 8-fold speedup for nucleotide-based models. As a practical demonstration, we also estimate the timing of the first introduction of West Nile virus into the continental Unites States under a codon model with a relaxed molecular clock from 104 full viral genomes, an inference task previously intractable. We provide an implementation of our GPU algorithms in BEAGLE v4.0.0, an open source library for statistical phylogenetics that enables parallel calculations on multi-core CPUs and GPUs.

Distinct antibody responses to endemic coronaviruses pre- and post-SARS-CoV-2 infection in Kenyan infants and mothers

Pre-existing antibodies that bind endemic human coronaviruses (eHCoVs) can cross-react with SARS-CoV-2, the betacoronavirus that causes COVID-19, but whether these responses influence SARS-CoV-2 infection is still under investigation and is particularly understudied in infants. In this study, we measured eHCoV and SARS-CoV-1 IgG antibody titers before and after SARS-CoV-2 seroconversion in a cohort of Kenyan women and their infants. Pre-existing eHCoV antibody binding titers were not consistently associated with SARS-CoV-2 seroconversion in infants or mothers, though we observed a very modest association between pre-existing HCoV-229E antibody levels and lack of SARS-CoV-2 seroconversion in infants. After seroconversion to SARS-CoV-2, antibody binding titers to endemic betacoronaviruses HCoV-OC43 and HCoV-HKU1, and the highly pathogenic betacoronavirus SARS-CoV-1, but not endemic alphacoronaviruses HCoV-229E and HCoV-NL63, increased in mothers. However, eHCoV antibody levels did not increase following SARS-CoV-2 seroconversion in infants, suggesting the increase seen in mothers was not simply due to cross-reactivity to naively generated SARS-CoV-2 antibodies. In contrast, the levels of antibodies that could bind SARS-CoV-1 increased after SARS-CoV-2 seroconversion in both mothers and infants, both of whom are unlikely to have had a prior SARS-CoV-1 infection, supporting prior findings that SARS-CoV-2 responses cross-react with SARS-CoV-1. In summary, we find evidence for increased eHCoV antibody levels following SARS-CoV-2 seroconversion in mothers but not infants, suggesting eHCoV responses can be boosted by SARS-CoV-2 infection when a prior memory response has been established, and that pre-existing cross-reactive antibodies are not strongly associated with SARS-CoV-2 infection risk in mothers or infants.

Macaque-human differences in SARS-CoV-2 Spike antibody response elicited by vaccination or infection

Macaques are a commonly used model for studying immunity to human viruses, including for studies of SARS-CoV-2 infection and vaccination. However, it is unknown whether macaque antibody responses recapitulate, and thus appropriately model, the response in humans. To answer this question, we employed a phage-based deep mutational scanning approach (Phage- DMS) to compare which linear epitopes are targeted on the SARS-CoV-2 Spike protein in humans and macaques following either vaccination or infection. We also used Phage-DMS to determine antibody escape pathways within each epitope, enabling a granular comparison of antibody binding specificities at the locus level. Overall, we identified some common epitope targets in both macaques and humans, including in the fusion peptide (FP) and stem helix- heptad repeat 2 (SH-H) regions. Differences between groups included a response to epitopes in the N-terminal domain (NTD) and C-terminal domain (CTD) in vaccinated humans but not vaccinated macaques, as well as recognition of a CTD epitope and epitopes flanking the FP in convalescent macaques but not convalescent humans. There was also considerable variability in the escape pathways among individuals within each group. Sera from convalescent macaques showed the least variability in escape overall and converged on a common response with vaccinated humans in the SH-H epitope region, suggesting highly similar antibodies were elicited. Collectively, these findings suggest that the antibody response to SARS-CoV-2 in macaques shares many features with humans, but with substantial differences in the recognition of certain epitopes and considerable individual variability in antibody escape profiles, suggesting a diverse repertoire of antibodies that can respond to major epitopes in both humans and macaques. Author summaryNon-human primates, including macaques, are considered the best animal model for studying infectious diseases that infect humans. Vaccine candidates for SARS-CoV-2 are first tested in macaques to assess immune responses prior to advancing to human trials, and macaques are also used to model the human immune response to SARS-CoV-2 infection. However, there may be differences in how macaque and human antibodies recognize the SARS-CoV-2 entry protein, Spike. Here we characterized the locations on Spike that are recognized by antibodies from vaccinated or infected macaques and humans. We also made mutations to the viral sequence and assessed how these affected antibody binding, enabling a comparison of antibody binding requirements between macaques and humans at a very precise level. We found that macaques and humans share some responses, but also recognize distinct regions of Spike. We also found that in general, antibodies from different individuals had unique responses to viral mutations, regardless of species. These results will yield a better understanding of how macaque data can be used to inform human immunity to SARS-CoV-2.

SARS-CoV-2 antibody binding and neutralization in dried blood spot eluates and paired plasma

Widescale assessment of SARS-CoV-2-specific antibodies is critical to understanding population seroprevalence, correlates of protection, and the longevity of vaccine-elicited responses. Most SARS-CoV-2 studies characterize antibody responses in plasma/sera. While reliable and broadly used, these samples pose several logistical restrictions such as requiring venipuncture for collection and cold chain for transportation and storage. Dried blood spots (DBS) overcome these barriers as they can be self-collected by fingerstick and mailed and stored at ambient temperature. Here, we evaluate the suitability of DBS for SARS-CoV-2 antibody assays by comparing several antibody responses between paired plasma and DBS from SARS-CoV-2 convalescent and vaccinated individuals. We found that DBS not only reflected plasma antibody binding by ELISA and epitope profiles using phage-display, but also yielded SARS-CoV-2 neutralization titers that highly correlated with paired plasma. Neutralization measurement was further streamlined by adapting assays to a high-throughput 384-well format. This study supports the adoption of DBS for numerous SARS-CoV-2 binding and neutralization assays.

High resolution profiling of pathways of escape for SARS-CoV-2 spike-binding antibodies

Defining long-term protective immunity to SARS-CoV-2 is one of the most pressing questions of our time and will require a detailed understanding of potential ways this virus can evolve to escape immune protection. Immune protection will most likely be mediated by antibodies that bind to the viral entry protein, Spike (S). Here we used Phage-DMS, an approach that comprehensively interrogates the effect of all possible mutations on binding to a protein of interest, to define the profile of antibody escape to the SARS-CoV-2 S protein using COVID-19 convalescent plasma. Antibody binding was common in two regions: the fusion peptide and linker region upstream of the heptad repeat region 2. However, escape mutations were variable within these immunodominant regions. There was also individual variation in less commonly targeted epitopes. This study provides a granular view of potential antibody escape pathways and suggests there will be individual variation in antibody-mediated virus evolution.

Epitope profiling reveals binding signatures of SARS-CoV-2 immune response and cross-reactivity with endemic HCoVs

A major goal of current SARS-CoV-2 vaccine efforts is to elicit antibody responses that confer protection. Mapping the epitope targets of the SARS-CoV-2 antibody response is critical for innovative vaccine design, diagnostics, and development of therapeutics. Here, we developed a phage display library to map antibody binding sites at high resolution within the complete viral proteomes of all human-infecting coronaviruses in patients with mild or moderate/severe COVID-19. The dominant immune responses to SARS-CoV-2 were targeted to regions spanning the Spike protein, Nucleocapsid, and ORF1ab. Some epitopes were identified in the majority of samples while others were rare, and we found variation in the number of epitopes targeted by different individuals. We also identified a set of cross-reactive sequences that were bound by antibodies in SARS-CoV-2 unexposed individuals. Finally, we uncovered a subset of enriched epitopes from commonly circulating human coronaviruses with significant homology to highly reactive SARS-CoV-2 sequences.