Biomedical subjects

Jack Sullivan

Publications and source records attributed to Jack Sullivan.

10 recordsLinked to original sources

PubMed2005

Evaluating the performance of a successive-approximations approach to parameter optimization in maximum-likelihood phylogeny estimation.

Almost all studies that estimate phylogenies from DNA sequence data under the maximum-likelihood (ML) criterion employ an approximate approach. Most commonly, model parameters are estimated on some initial phylogenetic estimate derived using a rapid method (neighbor-joining or parsimony). Parameters are then held constant during a tree search, and ideally, the procedure is repeated until convergence is achieved. However, the effectiveness of this approximation has not been formally assessed, in part because doing so requires computationally intensive, full-optimization analyses. Here, we report both indirect and direct evaluations of the effectiveness of successive approximations. We obtained an indirect evaluation by comparing the results of replicate runs on real data that use random trees to provide initial parameter estimates. For six real data sets taken from the literature, all replicate iterative searches converged to the same joint estimates of topology and model parameters, suggesting that the approximation is not starting-point dependent, as long as the heuristic searches of tree space are rigorous. We conducted a more direct assessment using simulations in which we compared the accuracy of phylogenies estimated using full optimization of all model parameters on each tree evaluated to the accuracy of trees estimated via successive approximations. There is no significant difference between the accuracy of the approximation searches relative to full-optimization searches. Our results demonstrate that successive approximation is reliable and provide reassurance that this much faster approach is safe to use for ML estimation of topology.

Algorithms↗

PubMed2005

Testing hypotheses of speciation timing in Dicamptodon copei and Dicamptodon aterrimus (Caudata: Dicamptodontidae).

Giant salamanders of the genus Dicamptodon are members of the mesic forest ecosystem that occurs in the Pacific Northwest of North America. We estimate the phylogeny of the genus to test several hypotheses concerning speciation and the origin of current species distributions. Specifically, we test competing a priori hypotheses of dispersal and vicariance to explain the disjunct inland distribution of the Idaho giant salamander (D. aterrimus) and to test the hypothesis of Pleistocene speciation of Cope's giant salamander (D. copei) using Bayesian hypothesis testing. We determined that available outgroups were too divergent to root the phylogeny effectively, and we calculated Bayesian posterior probabilities for each of the 15 possible root placements for this four-taxon group. This analysis placed the root on the branch leading to D. aterrimus, indicating that current distribution and speciation of D. aterrimus fit the ancient vicariance hypothesis and are attributable to the orogeny of the Cascade Mountains rather than recent inland dispersal. Furthermore, test results indicate that D. copei is distantly related to other coastal lineages and likely originated much earlier than the Pleistocene. These results suggest that speciation within the genus is attributable to ancient geologic events, while more recent Pleistocene glaciation has shaped genetic variation and distributions within the extant species.

Animals↗

PubMed2005

Maximum-likelihood methods for phylogeny estimation.

Maximum-likelihood (ML) estimation of phylogenies has reached a rather high level of sophistication because of algorithmic advances, improvements in models of sequence evolution, and improvements in statistical approaches and application of cluster computing. Here, I provide a brief basic background in application of the general principle of ML estimation to phylogenetics and provide an example of selecting among a nested set of ML models using a dynamic approach to hierarchical likelihood-ratio tests. I focus attention on PAUP* because it provides unique ease of switching among alternative optimality criteria (e.g., minimum evolution, parsimony, and ML). Further, examples of parametric bootstrap tests are provided that demonstrate statistical tests of phylogenetic hypotheses and model adequacy, in an absolute rather than relative sense. The increasing availability of clustered, parallelized computation makes use of such parametric approaches feasible.

Algorithms↗

PubMed2004

Accounting for uncertainty in the tree topology has little effect on the decision-theoretic approach to model selection in phylogeny estimation.

Currently available methods for model selection used in phylogenetic analysis are based on an initial fixed-tree topology. Once a model is picked based on this topology, a rigorous search of the tree space is run under that model to find the maximum-likelihood estimate of the tree (topology and branch lengths) and the maximum-likelihood estimates of the model parameters. In this paper, we propose two extensions to the decision-theoretic (DT) approach that relax the fixed-topology restriction. We also relax the fixed-topology restriction for the Bayesian information criterion (BIC) and the Akaike information criterion (AIC) methods. We compare the performance of the different methods (the relaxed, restricted, and the likelihood-ratio test [LRT]) using simulated data. This comparison is done by evaluating the relative complexity of the models resulting from each method and by comparing the performance of the chosen models in estimating the true tree. We also compare the methods relative to one another by measuring the closeness of the estimated trees corresponding to the different chosen models under these methods. We show that varying the topology does not have a major impact on model choice. We also show that the outcome of the two proposed extensions is identical and is comparable to that of the BIC, Extended-BIC, and DT. Hence, using the simpler methods in choosing a model for analyzing the data is more computationally feasible, with results comparable to the more computationally intensive methods. Another outcome of this study is that earlier conclusions about the DT approach are reinforced. That is, LRT, Extended-AIC, and AIC result in more complicated models that do not contribute to the performance of the phylogenetic inference, yet cause a significant increase in the time required for data analysis.

Computational Biology↗

PubMed2004

Testing nested phylogenetic and phylogeographic hypotheses in the Plethodon vandykei species group.

Mesic forests in the North American Pacific Northwest occur in two disjunct areas: along the coastal and Cascade ranges of Oregon, Washington, and British Columbia as well as the Northern Rocky Mountains of Idaho, Montana, and British Columbia. Over 150 species or species complexes have disjunct populations in each area, and a priori hypotheses based on phytogeography and geology potentially explain the disjunction via either dispersal or vicariance. Here, we test these hypotheses in the disjunct salamander complex Plethodon vandykei and P. idahoensisby collecting genetic data (669 bp of Cyt b) from 262 individuals. Maximum likelihood analysis indicated reciprocal monophyly of these species, supporting the ancient vicariance hypothesis, whereas parametric bootstrap and Bayesian hypothesis testing allow rejection of the dispersal hypothesis. The coalescent estimate of the time since population divergence (estimated using MDIV) is 3.75 x 106 years, and the 95%credibility interval of this value overlaps with the geological estimate of vicariance, but not the hypothesized dispersal. These results are congruent with the pattern seen in other mesic forest amphibian lineages and suggest disjunction in amphibians may be a concerted response to a geological/climatological event. WithinP. idahoensis, we tested the corollary hypothesis of an inland Pleistocene refugium in the Clearwater drainage with nested clade analysis and coalescent estimates of population growth rate (g). Both analyses support post-Pleistocene expansion from the Clearwater refugium. We corroborated this result by calculating Tajima's Dand mismatch distribution within each drainage, showing strong evidence for recent population expansion within most drainages. This work demonstrates the utility of statistical phylogeography and contributes two novel analytical tools: tests of stationarity with respect to topology in the Bayesian estimation, and the use of coalescent simulations to test the significance of the population growth-rate parameter.

Analysis of Variance↗

PubMed2003

Extensive mtDNA variation within the yellow-pine chipmunk, Tamias amoenus (Rodentia: Sciuridae), and phylogeographic inferences for northwest North America.

The yellow-pine chipmunk, Tamias amoenus, is common in xerophytic forests throughout much of northwest North America. We analyzed cytochrome b sequence variation from 155 individuals representing 57 localities across the distribution of T. amoenus including 10 additional species of Tamias. Maximum likelihood and parsimony tree estimation methods were used in conjunction with nested clade analysis to infer both deep and population-level processes. Our results indicate that two currently recognized subspecies of T. amoenus (T. a. canicaudus and T. a. cratericus) are not nested within other samples of T. amoenus. Maximum uncorrected levels of intraspecific sequence divergence within remaining samples of T. amoenus are >7%. Substantial geographic variation is characterized by 12 well-supported clades that correspond to distinct mountain ranges, but do not necessarily follow existing subspecific taxonomy. Significant association between geography and genealogy was detected within many of these clades and can be attributed to different population-level processes including past fragmentation, recent range expansion, and isolation by distance.

Animals↗

PubMed2003

Performance-based selection of likelihood models for phylogeny estimation.

Phylogenetic estimation has largely come to rely on explicitly model-based methods. This approach requires that a model be chosen and that that choice be justified. To date, justification has largely been accomplished through use of likelihood-ratio tests (LRTs) to assess the relative fit of a nested series of reversible models. While this approach certainly represents an important advance over arbitrary model selection, the best fit of a series of models may not always provide the most reliable phylogenetic estimates for finite real data sets, where all available models are surely incorrect. Here, we develop a novel approach to model selection, which is based on the Bayesian information criterion, but incorporates relative branch-length error as a performance measure in a decision theory (DT) framework. This DT method includes a penalty for overfitting, is applicable prior to running extensive analyses, and simultaneously compares all models being considered and thus does not rely on a series of pairwise comparisons of models to traverse model space. We evaluate this method by examining four real data sets and by using those data sets to define simulation conditions. In the real data sets, the DT method selects the same or simpler models than conventional LRTs. In order to lend generality to the simulations, codon-based models (with parameters estimated from the real data sets) were used to generate simulated data sets, which are therefore more complex than any of the models we evaluate. On average, the DT method selects models that are simpler than those chosen by conventional LRTs. Nevertheless, these simpler models provide estimates of branch lengths that are more accurate both in terms of relative error and absolute error than those derived using the more complex (yet still wrong) models chosen by conventional LRTs. This method is available in a program called DT-ModSel.

Bayes Theorem↗

PubMed2003

Phylogeography and introgressive hybridization: chipmunks (genus Tamias) in the northern Rocky Mountains.

If phylogeographic studies are to be broadly used for assessing population-level processes relevant to speciation and systematics, the ability to identify and incorporate instances of hybridization into the analytical framework is essential. Here, we examine the evolutionary history of two chipmunk species, Tamias ruficaudus and Tamias amoenus, in the northern Rocky Mountains by integrating multivariate morphometrics of bacular (os penis) variation, phylogenetic estimation, and nested clade analysis with regional biogeography. Our results indicate multiple examples of mitochondrial DNA introgression layered within the evolutionary history of these nonsister species. Three of these events are most consistent with recent and/or ongoing asymmetric introgression of mitochondrial DNA across morphologically defined secondary contact zones. In addition, we find preliminary evidence where a fourth instance of nonconcordant characters may represent complete fixation of introgressed mitochondrial DNA via a more ancient hybridization event, although alternative explanations of convergence or incomplete sorting of ancestral polymorphisms cannot be dismissed with these data. The demonstration of hybridization among chipmunks with strongly differentiated bacular morphology contradicts long-standing assumptions that variation within this character is diagnostic of complete reproductive isolation within Tamias. Our results illustrate the utility of phylogeographic analyses for detecting instances of reticulate evolution and for incorporating this and other information in the inference of the evolutionary history of species.

Animals↗

PubMed2002

Evaluating hypotheses of deuterostome phylogeny and chordate evolution with new LSU and SSU ribosomal DNA data.

We investigated evolutionary relationships among deuterostome subgroups by obtaining nearly complete large-subunit ribosomal RNA (LSU rRNA)-gene sequences for 14 deuterostomes and 3 protostomes and complete small-subunit (SSU) rRNA-gene sequences for five of these animals. With the addition of previously published sequences, we compared 28 taxa using three different data sets (LSU only, SSU only, and combined LSU + SSU) under minimum evolution (with LogDet distances), maximum likelihood, and maximum parsimony optimality criteria. Additionally, we analyzed the combined LSU + SSU sequences with spectral analysis of LogDet distances, a technique that measures the amount of support and conflict within the data for every possible grouping of taxa. Overall, we found that (1) the LSU genes produced a tree very similar to the SSU gene tree, (2) adding LSU to SSU sequences strengthened the bootstrap support for many groups above the SSU-only values (e.g., hemichordates plus echinoderms as Ambulacraria; lancelets as the sister group to vertebrates), (3) LSU sequences did not support SSU-based hypotheses of pterobranchs evolving from enteropneusts and thaliaceans evolving from ascidians, and (4) the combined LSU + SSU data are ambiguous about the monophyly of chordates. No tree-building algorithm united urochordates conclusively with other chordates, although spectral analysis did so, providing our only evidence for chordate monophyly. With spectral analysis, we also evaluated several major hypotheses of deuterostome phylogeny that were constructed from morphological, embryological, and paleontological evidence. Our rRNA-gene analysis refutes most of these hypotheses and thus advocates a rethinking of chordate and vertebrate origins.

Animals↗

PubMed2000

Comparative Phylogeography of Mesoamerican Highland Rodents: Concerted versus Independent Response to Past Climatic Fluctuations.

The phylogeography of Sumichrast's harvest mouse (Reithrodontomys sumichrasti) was examined through maximum-likelihood and parsimony analyses of 1,130 bp of mitochondrial Cytochrome b sequence data from 43 individuals. The phylogeography of this Middle American highland forest-dwelling species was compared to that previously published for the codistributed Aztec deer mouse complex (Peromyscus aztecus/Peromyscus hylocetes complex) in order to test competing hypotheses of concerted versus independent responses of codistributed forms to past climatic fluctuations. Qualitatively, there were strong similarities in the phylogeographic patterns of the two groups, yet there were also areas of incongruence. Likelihood-ratio tests (Kishino-Hasegawa-Templeton and parametric bootstrap tests) indicated that this incongruence is significant and cannot be attributed simply to uncertainty in phylogenetic estimation, thereby falsifying the concerted-response hypothesis. Conversely, tree-reconciliation analysis of the area relationships inferred for each group separately indicated that there has been a significant history of covicariance between the two groups, falsifying the independent-response hypothesis. It appears that codistributed taxa in the geologically complex highlands of Mesoamerica share more common biogeographical history than can be accounted for by the independent-response hypothesis yet have not responded to past climatic fluctuations in the lock-step fashion predicted by the concerted-response hypothesis.

Middle American highlands↗