Biology subjects

Mitov, V.

Publications and source records attributed to Mitov, V..

3 recordsLinked to original sources

bioRxiv2017

Fast Bayesian Inference of Phylogenetic Models Using Parallel Likelihood Calculation and Adaptive Metropolis Sampling

O_LIPhylogenetic comparative models (PCMs) have been used to study macroevolutionary patterns, to characterize adaptive phenotypic landscapes, to quantify rates of evolution, to measure the heritability of traits, and to test various evolutionary hypotheses. A major obstacle to applying these models has been the complexity of evaluating their likelihood function. Recent works have shown that for many PCMs, the likelihood can be obtained in time proportional to the size of the tree based on post-order tree traversal, also known as pruning. Despite this progress, inferring complex multi-trait PCMs on large trees remains a time-intensive task. Here, we study parallelizing the pruning algorithm as a generic technique for speeding-up PCM-inference.\nC_LIO_LIWe implement several parallel traversal algorithms in the form of a generic C++ library for Serial and Parallel LIneage Traversal of Trees (SPLITT). Based on SPLITT, we provide examples of parallel likelihood evaluation for several popular PCMs, ranging from a single-trait Brownian motion model to complex multi-trait Ornstein-Uhlenbeck and mixed Gaussian phylogenetic models.\nC_LIO_LIUsing the phylogenetic Ornstein-Uhlenbeck mixed model (POUMM) as a showcase, we run benchmarks on up to 24 CPU cores, reporting up to an order of magnitude parallel speed-up on simulated balanced and unbalanced trees of up to 100,000 tips with up to 16 traits. Noticing that the parallel speed-up depends on multiple factors, the SPLITT library is capable to automatically select the fastest traversal strategy for a given hardware, tree-topology, and data. Combining SPLITT likelihood calculation with adaptive Metropolis sampling on real data, we show that the time for Bayesian POUMM inference on a tree of 10,000 tips can be reduced from several days to minutes.\nC_LIO_LIWe conclude that parallel pruning effectively accelerates the likelihood calculation and, thus, the statistical inference of Gaussian phylogenetic models. For time-intensive Bayesian inferences, we recommend combining this technique with adaptive Metropolis sampling. Beyond Gaussian models, the parallel tree traversal can be applied to numerous other models, including discrete trait and birth-death population dynamics models. Currently, SPLITT supports multi-core shared memory architectures, but can be extended to distributed memory architectures as well as graphical processing units.\nC_LI

evolutionary biology↗

bioRxiv2017

Dissecting HIV Virulence: Heritability Of Setpoint Viral Load, CD4+ T Cell Decline And Per-Parasite Pathogenicity

Pathogen strains may differ in virulence because they attain different loads in their hosts, or because they induce different disease-causing mechanisms independent of their load. In evolutionary ecology, the latter is referred to as \"per-parasite pathogenicity\". Using viral load and CD4+ T cell measures from 2014 HIV-1 subtype B infected individuals enrolled in the Swiss HIV Cohort Study, we investigated if virulence -- measured as the rate of decline of CD4+ T cells -- and per-parasite pathogenicity are heritable from donor to recipient. We estimated heritability by donor-recipient regressions applied to 196 previously identified transmission pairs, and by phylogenetic mixed models applied to a phylogenetic tree inferred from HIV pol sequences. Regressing the CD4+ T cell declines and per-parasite pathogenicities of the transmission pairs did not yield heritability estimates significantly different from zero. With the phylogenetic mixed model, however, our best estimate for the heritability of the CD4+ T cell decline is 17% (5%-30%), and that of the per-parasite pathogenicity is 17% (4%-29%). Further, we confirm that the set-point viral load is heritable, and estimate a heritability of 29% (12%-46%). Interestingly, the pattern of evolution of all these traits differs significantly from neutrality, and is most consistent with stabilizing selection for the set-point viral load, and with directional selection for the CD4+ T cell decline and the per-parasite pathogenicity. Our analysis shows that the viral genetype affects virulence mainly by modulating the per-parasite pathogenicity, while the indirect effect via the set-point viral load is minor.

evolutionary biology↗

bioRxiv2017

POUMM: An R-package for Bayesian Inference of Phylogenetic Heritability

AO_SCPLOWBSTRACTC_SCPLOWPhylogenetic comparative methods have been used to model trait evolution, to test selection versus neutral hypotheses, to estimate optimal trait-values, and to quantify the rate of adaptation towards these optima. Several authors have proposed algorithms calculating the likelihood for trait evolution models, such as the Ornstein-Uhlenbeck (OU) process, in time proportional to the number of tips in the tree. Combined with gradient-based optimization, these algorithms enable maximum likelihood (ML) inference within seconds, even for trees exceeding 10,000 tips. Despite its useful statistical properties, ML has been criticised for being a point estimator prone to getting stuck in local optima. As an elegant alternative, Bayesian inference explores the entire information in the data and compares it to prior knowledge but, usually, runs in much longer time, even for small trees. Here, we propose an approach to use the full potential of ML and Bayesian inference, while keeping the runtime within minutes. Our approach combines (i) a new algorithm for parallel likelihood calculation; (ii) a previously published method for adaptive Metropolis sampling. In principle, the strategy of (i) and (ii) can be applied to any likelihood calculation on a tree which proceeds in a pruning-like fashion leading to enormous speed improvements. As a showcase, we implement the phylogenetic Ornstein-Uhlenbeck mixed model (POUMM) in the form of an easy-to-use and highly configurable R-package. In addition to the above-mentioned usage of comparative methods, the POUMM allows to estimate non-heritable variance and phylogenetic heritability. Using simulations and empirical data from 487 mammal species, we show that the POUMM is far more reliable in terms of unbiased estimates and false positive rate for stabilizing selection, compared to its alternative - the non-mixed Ornstein-Uhlenbeck model, which assumes a fully heritable and perfectly measurable trait. Further, our analysis reveals that the phylogenetic mixed model (PMM), which assumes neutral evolution (Brownian motion) can be a very unstable estimator of phylogenetic heritability, even if the Brownian motion assumption is only weakly violated. Our results prove the need for a simultaneous account for selection and non-heritable variance in phylogenetic evolutionary models and challenge stabilizing selection hypotheses stated in numerous macro-evolutionary studies.

evolutionary biology↗