Maximum Composite Likelihood Research Articles

A count data model with endogenous covariates: Formulation and application to roadway crash frequency at intersections

This paper proposes an estimation approach for count data models with endogenous covariates. The maximum approximate composite marginal likelihood inference approach is used to estimate model parameters. The modeling framework is applied to predict crash frequency at urban intersections in Irving, Texas. The sample is drawn from the Texas Department of Transportation (TxDOT) crash incident files for the year 2008. The results highlight the importance of accommodating endogeneity effects in count models. In addition, the results reveal the increased propensity for crashes at intersections with flashing lights, intersections with crest approaches, and intersections that are on frontage roads.

Nucleotide based validation of Ocimum species by evaluating three candidate barcodes of the chloroplast region

The genus Ocimum comprises of several medicinally important species which frequently fall prey to adulteration due to misidentification. A proficient method is hence required to solve the problems that exist in differentiating its various morphotypes. In plants, candidate DNA barcodes of the chloroplast and nuclear regions have proved to be a great success in the validation of several plant families. Hence, this study involves the use of the molecular-based DNA barcoding method to identify some of the most common and useful species of the genus Ocimum (Tulsi). Here, DNA amplification of three candidate barcodes of the chloroplast genome viz. matK, rbcL and psbA-trnH was performed, to access their ability to produce high sequence variability. The discrimination among species was performed using the Kimura 2-parameter and maximum composite likelihood methods. On analysing the sequence data, the psbA-trnH region proved to be the most suitable candidate barcode and gave an overall variation of 7.3% at the interspecies level. A clear differentiation was found at the species level, showing a maximum distance of 0.264 between dissimilar species. Also, phylogenetic analysis led to the successful identification of hybrids, while it failed to do so at the variety level. Hence, it can be inferred that DNA barcoding is ideal for species-level identification of the genus Ocimum.

Hepatitis B virus subgenotype A1, occurrence of subgenotype D4, and S gene mutations among voluntary blood donors in Kenya

Kenya is one of the high endemic zones for hepatitis B virus (HBV) infection. The consensuses on prevalence of the HBV genotypes and the existence of their variants have not been fully established in Kenya. Hence, there is a need to further monitor the diversity of HBV. This study aimed to extend the current molecular and epidemiological information about the geographical distribution of HBV genotypes and subgenotypes, as well as to describe the hepatitis B surface antigen (HBsAg) variants circulating in different Regional Blood Transfusion Centres of Kenya. A total of 32 HBsAg positive blood units from five different blood transfusion centers in Kenya were used in the study. The HBV DNA preS/S-gene was amplified and sequenced. Alignments of S gene were applied using reference sequence from GeneBank. Phylogenetic analysis was performed using the MEGAv4.0 software with the neighbor-joining and maximum composite likelihood methods. Twenty-one plasma samples (65.6%) were DNA positive and were successfully sequenced. Eighteen out of the twenty-one isolates (85.7%) belonged to subgenotype A1 Afro-Asian: six were from Nairobi, four from Kisumu, two from Embu, and three each from Eldoret and Mombasa. The other three strains (14.3%, 3/21) belonged to subgenotype D4 from Mombasa. The HBsAg mutations were detected in nine isolates (42.9%, 9/21). The HBV/A1 and HBV/D4 are dominant among blood donors in Kenya. This demonstrates that continuous monitoring of the HBV diversity would help reveal circulating genotypes and subgenotypes as well as mutants of clinical significance in Kenya.

BartonellaSpecies in Raccoons and Feral Cats, Georgia, USA

<i>Bartonella</i>Species in Raccoons and Feral Cats, Georgia, USA

Genome-Scale Approach and the Performance of Phylogenetic Methods

The use of genome-scale approach in phylogenetic analysis is imperative in order to resolve evolutionary relationships over large taxon sets and deep phylogenetic divergences. But yet it is not clear what are the strengths and weaknesses of the various phylogenetic methods or which one should be preferred under genome-scale approach. In this article, the performance of five major phylogenetic methods is evaluated under genome-scale approach using biologically realistic simulated data. The following phylogenetic methods are considered; Bayesian, maximum likelihood (ML), neighbor joining (NJ), NJ maximum composite likelihood (NJ-MCL), and maximum parsimony (MP). Simulation results show that probabilistic methods (i.e., Bayesian and ML methods) are much more accurate than the NJ-MCL, MP and NJ methods. Concerning the consistency of methods, ML is consistent than other methods. This analysis shows that the NJ-MCL, MP, and NJ methods are fast (i.e., computationally efficient), but their accuracy and consistency are very poor compared to Bayesian and ML methods. On the other hand, the Bayesian method is an accurate one, but less consistent than the ML method, and it takes much longer execution time. Therefore, based on the accuracy, consistency and computational efficiency the ML method is the preferred algorithm under genome-scale approach. In addition to the methods performance, this study has investigated several important aspects of genomescale phylogeny; such as how concatenations of longest and smallest genes make effect on the method’s performance, how much datasets are needed to recover the true tree (i.e. true evolutionary history of a group of species or genes), and whether more genes or more characters are important. These are explained in the result section.

Detection of Bovine Leukemia Virus in Brains of Cattle with a Neurological Syndrome: Pathological and Molecular Studies

Bovine leukemia virus (BLV) was investigated in the central nervous system (CNS) of cattle with neurological syndrome. A total of 269 CNS samples were submitted to nested-PCR (BLV env gene gp51), and the viral genotypes were identified. The nested-PCR was positive in 4.8% (13/269) CNS samples, with 2.7% (2/74) presenting at histological examination lesions of nonpurulent meningoencephalitis (NPME), whereas 5.6% (11/195) not presenting NPME (P > 0.05). No samples presented lymphosarcoma. The PCR products (437 bp) were sequenced and submitted to phylogenetic analysis by neighbor-joining and maximum composite likelihood methods, and genotypes 1, 5, and 6 were detected, corroborating other South American studies. The genotype 6 barely described in Brazil and Argentina was more frequently detected in this study. The identity matrices showed maximum similarity (100%) among some samples of this study and one from Argentina (FJ808582), recovered from GenBank. There was no association among the genotypes and NPME lesions.

Molecular and phylogenetic analysis of matrix gene of avian influenza viruses isolated from wild birds and live bird markets in the USA

Please cite this paper as: Chander et al. (2012) Molecular and phylogenetic analysis of matrix gene of avian influenza viruses isolated from wild birds and live bird markets in the USA. Influenza and Other Respiratory Viruses 7(4), 513–520. Background Wild birds are the natural hosts for influenza A viruses (IAVs) and provide a niche for the maintenance of this virus. Objectives This study was undertaken to analyze nucleotide sequences of the matrix (M) gene of AIVs isolated from wild birds and live bird markets (LBMs) to index the changes occurring in this gene. Methods M‐gene of 229 avian influenza virus (AIV) isolates obtained from wild birds and LBMs was amplified and sequenced. Full‐length sequences (∼900 nt.) thus obtained were analyzed to identify changes that may be associated with resistance to adamantanes. Phylogenetic analysis of all sequences was performed using clustalw, and evolutionary distances were calculated by maximum composite likelihood method using mega (ver. 5.0) software. Results Twenty‐seven different viral subtypes were represented with H3N8 being the most dominant subtype in wild birds and H7N2 being the predominant subtype among isolates from LBMs. Phylogenetic analysis of the M‐gene showed a high degree of nucleotide sequence identity with US isolates of AIVs but not with those of Asian or European lineages. While none of the isolates from wild birds had any antiviral resistance–associated mutations, 17 LBM isolates carried polymorphisms known to cause reduced susceptibility to antiviral drugs (adamantanes). Of these 17 isolates, 16 had S31N change and one isolate had V27A mutation. Conclusions These results indicate independent evolution of M‐gene in the absence of any antiviral drugs leading to mutations causing resistance indicating the need for continued active surveillance of AIVs.

Molecular characterization of subcutaneous and muscular coenurosis of goats in United Arab Emirates

Coenurosis is a parasitic disease caused by the larval stage of Taenia multiceps, which affects various ruminants species, particularly sheep and goats, and occasionally humans. In this note, non-cerebral coenurosis in goats with the goal to compare morphological and biomolecular characteristics is focused in order to determine if the non-cerebral specimens should be considered a new specie or a strain of T. multiceps. Three hundred goats slaughtered with macroscopic lesions due to the presence of metacestodes were examined in an abattoir in Dubai, United Arab Emirates in order to evaluate the presence of extra-cerebral coenuri. Forty-eight coenuri were found under the skin, between fasciae of the skeletal muscles, diaphragm and in the abdominal cavities. Morphologically, the examined non-cerebral coenuri recovered showed the same features reported by other authors for Coenurus cerebralis, but their location outside the central nervous system suggests that they may be a different strain or genetic variants of T. multiceps. Nine caprine coenuri were then processed for sequencing of mitochondrial partial COI (396 bp) and ND1 (471 bp) genes indicating that they had a pairwise distance of 1.0–1.3% and 2.4–4.1% compared with parasite’ COI sequences from Italy (Tm1–Tm3 strains) and Erzurum strains of ovine origin, respectively; whereas it had 0.6–1.3% and 0.4–1.1% pairwise distance for ND1. Phylogenetic trees of their ND1 and COI sequences using the maximum composite likelihood method with MEGA showed that these nine parasites recovered in Emirates abattoir could be grouped into a new strain. In conclusion, morphological and molecular characterization of caprine coenurosis from Dubai, United Arab Emirates, shows that the specimens are significantly different from other strains or genotypes in COI sequence, which suggests that they should belong to different genotypes or strains of T. multiceps.

Composite Partial Likelihood Estimation Under Length-Biased Sampling, With Application to a Prevalent Cohort Study of Dementia

The Canadian Study of Health and Aging (CSHA) employed a prevalent cohort design to study survival after onset of dementia, where patients with dementia were sampled and the onset time of dementia was determined retrospectively. The prevalent cohort sampling scheme favors individuals who survive longer. Thus, the observed survival times are subject to length bias. In recent years, there has been a rising interest in developing estimation procedures for prevalent cohort survival data that not only account for length bias but also actually exploit the incidence distribution of the disease to improve efficiency. This article considers semiparametric estimation of the Cox model for the time from dementia onset to death under a stationarity assumption with respect to the disease incidence. Under the stationarity condition, the semiparametric maximum likelihood estimation is expected to be fully efficient yet difficult to perform for statistical practitioners, as the likelihood depends on the baseline hazard function in a complicated way. Moreover, the asymptotic properties of the semiparametric maximum likelihood estimator are not well-studied. Motivated by the composite likelihood method (Besag 1974), we develop a composite partial likelihood method that retains the simplicity of the popular partial likelihood estimator and can be easily performed using standard statistical software. When applied to the CSHA data, the proposed method estimates a significant difference in survival between the vascular dementia group and the possible Alzheimer's disease group, while the partial likelihood method for left-truncated and right-censored data yields a greater standard error and a 95% confidence interval covering 0, thus highlighting the practical value of employing a more efficient methodology. To check the assumption of stable disease for the CSHA data, we also present new graphical and numerical tests in the article. The R code used to obtain the maximum composite partial likelihood estimator for the CSHA data is available in the online Supplementary Material, posted on the journal web site.

A new approach to specify and estimate non-normally mixed multinomial probit models

The current paper proposes the use of the multivariate skew-normal distribution function to accommodate non-normal mixing in cross-sectional and panel multinomial probit (MNP) models. The combination of skew-normal mixing and the MNP kernel lends itself nicely to estimation using Bhat’s (2011) maximum approximate composite marginal likelihood (MACML) approach. Simulation results for the cross-sectional case show that our proposed approach does well in recovering the underlying parameters, and also highlights the pitfalls of ignoring non-normality of the continuous mixing distribution when such non-normality is present. At the same time, the proposed model obviates the need to assume a pre-specified parametric distribution for the mixing, and allows the estimation of a very flexible, but still parsimonious, mixing distribution form.

Chloroplast Microsatellite Diversity Among and Within Prunus mahaleb L. and P. avium L. Species

Genetic diversity of 58 Mahaleb cherry (Prunus mahaleb) genotypes and six sweet cherry (P. avium) accessions was studied using 25 cpSSR primer pairs. Thirteen out of them demonstrated two to four alleles with an average of 2.46 alleles per primer pair. The mean of PIC value for the primers was 0.32. The average values of expected heterozygosity (He) and Shannon’s information index (I) for all loci were 0.35 and 0.55, respectively. The dendrogram based on cpSSR markers has been illustrated by MEGA4 software with Maximum Composite likelihood model and the Neighbor-joining method, which clustered the genotypes into four groups. Sum of first three principal components analysis (PCAs) could be represented most of (63.07%) the total variation in the original dimensions and confirmed the results of cluster analysis. Based on the AMOVA results, the allele numbers among groups and species studied in this research were more than those observed within them. In this study cpSSR markers provided a good tool for assessment of genetic diversity among and within mahaleb cherry and sweet cherry genotypes.

Bayesian inference from composite likelihoods, with an application to spatial extremes

Composite likelihoods are increasingly used in applications where the full likelihood is analytically unknown or computationally prohibitive. Although some frequentist properties of the maximum composite likelihood estimator are akin to those of the maximum likelihood estimator, Bayesian inference based on composite likelihoods is in its early stages. This paper discusses inference when one uses composite likelihood in Bayes' formula. We establish that using a composite likelihood results in a proper posterior density, though it can differ considerably from that stemming from the full likelihood. Building on previous work on composite likelihood ratio tests, we use asymptotic theory for misspecified models to propose two adjustments to the composite likelihood to obtain appropriate inference. We also investigate use of the Metropolis Hastings algorithm and two implementations of the Gibbs sampler for obtaining draws from the composite posterior. We test the methods on simulated data and apply them to a spatial extreme rainfall dataset. For the simulated data, we find that posterior credible intervals yield appropriate empirical coverage rates. For the extreme precipitation data, we are able to both effectively model marginal behavior throughout the study region and obtain appropriate measures of spatial dependence.

On the likelihood function of Gaussian max-stable processes

SUMMARY WederiveaclosedformexpressionforthelikelihoodfunctionofaGaussianmax-stableprocessindexed by R d at p d + 1 sites, d 1. We demonstrate the gain in efficiency in the maximum composite likelihood estimators of the covariance matrix from p = 2t op =3 sites in R 2 by means of a Monte Carlo

A simulation evaluation of the maximum approximate composite marginal likelihood (MACML) estimator for mixed multinomial probit models

This paper evaluates the ability of the maximum approximate composite marginal likelihood (MACML) estimation approach to recover parameters from finite samples in mixed cross-sectional and panel multinomial probit models. Comparisons with the maximum simulated likelihood (MSL) estimation approach are also undertaken. The results indicate that the MACML approach recovers parameters much more accurately than the MSL approach in all model structures and covariance specifications. The MACML inference approach also estimates the parameters efficiently, with the asymptotic standard errors being, in general, only a small proportion of the true values. As importantly, the MACML inference approach takes only a very small fraction of the time needed for MSL estimation. In particular, the results suggest that, for the case of five random coefficients, the MACML approach is about 50 times faster than the MSL for the cross-sectional random coefficients case, about 15 times faster than the MSL for the panel inter-individual random coefficients case, and about 350 times or more faster than the MSL for the panel intra- and inter-individual random coefficients case. As the number of alternatives in the unordered-response model increases, one can expect even higher computational efficiency factors for the MACML over the MSL approach. Further, as should be evident in the panel intra- and inter-individual random coefficients case, the MSL is all but practically infeasible when the mixing structure leads to an explosion in the dimensionality of integration in the likelihood function, but these situations are handled with ease in the MACML approach. It is hoped that the MACML procedure will spawn empirical research into rich model specifications within the context of unordered multinomial choice modeling, including autoregressive random coefficients, dynamics in coefficients, space–time effects, and spatial/social interactions.

The maximum approximate composite marginal likelihood (MACML) estimation of multinomial probit-based unordered response choice models

The likelihood functions of multinomial probit (MNP)-based choice models entail the evaluation of analytically-intractable integrals. As a result, such models are usually estimated using maximum simulated likelihood (MSL) techniques. Unfortunately, for many practical situations, the computational cost to ensure good asymptotic MSL estimator properties can be prohibitive and practically infeasible as the number of dimensions of integration rises. In this paper, we introduce a maximum approximate composite marginal likelihood (MACML) estimation approach for MNP models that can be applied using simple optimization software for likelihood estimation. It also represents a conceptually and pedagogically simpler procedure relative to simulation techniques, and has the advantage of substantial computational time efficiency relative to the MSL approach. The paper provides a “blueprint” for the MACML estimation for a wide variety of MNP models.

On the robustness of maximum composite likelihood estimate

Composite likelihood methods have been receiving growing interest in a number of different application areas, where the likelihood function is too cumbersome to be evaluated. In the present paper, some theoretical properties of the maximum composite likelihood estimate (MCLE) are investigated in more detail. Robustness of consistency of the MCLE is studied in a general setting, and clarified and illustrated through some simple examples. We also carry out a simulation study of the performance of the MCLE in a constructed model suggested by Arnold (2010) that is not multivariate normal, but has multivariate normal marginal distributions.

Genetic composition of contemporary swine influenza viruses in the West Central region of the United States of America.

Please cite this paper as: Evseenko et al. (2011) Genetic composition of contemporary swine influenza viruses in the West Central region of the United States of America. Influenza and Other Respiratory Viruses DOI: 10.1111/j.1750‐2659.2010.00189.x. Background Because of continuous circulation in different animal species and humans, influenza viruses have host‐specific phenotypic and genetic features. Reassortment of the genome segments can significantly change virus phenotype, potentially generating virus with pandemic potential. In 2009, a new pandemic influenza virus emerged. Objectives In this study, we attempted to find precursor viruses or genes of pandemic H1N1 influenza 2009 among 25 swine influenza viruses, isolated in the West Central region of the United States of America (USA), between 2007 and 2009. The Phylogenetically Similar Triple‐Reassortant Internal Genes (PSTRIG) cassette of all the viruses studied here as well as the PSTRIG cassette of pandemic H1N1 viruses have close but equidistant phylogenetic relationships to the early triple‐reassortant swine H3N2 influenza A isolated in the USA in 1998. Methods Samples (nasal swabs and lung tissue lavage) were taken from swine with or without clinical signs of respiratory disease via farmer‐funded syndromic surveillance. All studied viruses were isolated in Madin–Darby Canine Kidney cell cultures from the above‐mentioned samples according to standard protocols recommended for influenza virus isolation. Sequences were obtained using BigDye Terminator v3.1 Cycle Sequencing kit. Phylogenetic trees were built with MEGA 4.0 software using maximum composite likelihood algorithm and neighbor‐joining method for tree topology reconstruction. Results Among the 25 viruses studied, we have not found any gene segments of Eurasian origin. Our results suggest that pandemic H1N1 viruses diverged and are not directly descended from swine viruses that have been circulating in USA since 1998.

Multi locus sequence typing of Chlamydia reveals an association between Chlamydia psittaci genotypes and host species.

Chlamydia comprises a group of obligate intracellular bacterial parasites responsible for a variety of diseases in humans and animals, including several zoonoses. Chlamydia trachomatis causes diseases such as trachoma, urogenital infection and lymphogranuloma venereum with severe morbidity. Chlamydia pneumoniae is a common cause of community-acquired respiratory tract infections. Chlamydia psittaci, causing zoonotic pneumonia in humans, is usually hosted by birds, while Chlamydia abortus, causing abortion and fetal death in mammals, including humans, is mainly hosted by goats and sheep. We used multi-locus sequence typing to asses the population structure of Chlamydia. In total, 132 Chlamydia isolates were analyzed, including 60 C. trachomatis, 18 C. pneumoniae, 16 C. abortus, 34 C. psittaci and one of each of C. pecorum, C. caviae, C. muridarum and C. felis. Cluster analyses utilizing the Neighbour-Joining algorithm with the maximum composite likelihood model of concatenated sequences of 7 housekeeping fragments showed that C. psittaci 84/2334 isolated from a parrot grouped together with the C. abortus isolates from goats and sheep. Cluster analyses of the individual alleles showed that in all instances C. psittaci 84/2334 formed one group with C. abortus. Moving 84/2334 from the C. psittaci group to the C. abortus group resulted in a significant increase in the number of fixed differences and elimination of the number of shared mutations between C. psittaci and C. abortus. C. psittaci M56 from a muskrat branched separately from the main group of C. psittaci isolates. C. psittaci genotypes appeared to be associated with host species. The phylogentic tree of C. psittaci did not follow that of its host bird species, suggesting host species jumps. In conclusion, we report for the first time an association between C. psittaci genotypes with host species.

Sequencing of mitochondrial DNA cytochrome oxidase subunit I gene in sarcosaphagous flies from 14 provinces in China

To detect the 278 bp region of gene of the cytochrome oxidase subunit I (COI) in mitochondral DNA (mtDNA) of sarcosaphagous flies, identify the species of sarcosaphagous flies, and provide reference for forensic application. Samples were collected in Baotou and Chifeng of Inner Mongolia, Tianjin, Nanning, Fuzhou, Linyi of Shandong, Shijiazhuang, Yinchuan, Lanzhou, Huairou of Beijing, Xinxiang and Nanyang of Henan, Datong of Shanxi, Wuhu of Anhui, Quzhou of Zhejiang, Changsha, Zhuzhou and Yongzhou of Hunan. A total of 38 flies were randomly collected from rabbits, dogs and pigs which were set outdoors, then the flies' mitochondrial DNA (mtDNA) were extracted by the improved small insects DNA homogenate method. Amplification was conducted by Perkin-Elmer 9600 thermal cycler, then vertical non-denaturing 7% polyacrylamide gelectrophoresis. PCR products were purified using the nucleic acid purification kit. Sequences of both strands were obtained by direct sequence of the double-stranded PCR product using one of the PCR primers and the ABI PRISM big dye terminator cycle sequencing dit. Sequence reactions were electrophorsed on ABI Model 3730 DNA Sequencers. A UPGMA tree was contrasted using the maximum composite likelihood method in MEGA4. The 38 sarcosaphagous flies belonged to 3 families(Muscidae, Calliphoridae, and Sarcophagidae), 10 genuses (Musca Linnaeus, Hydrotaea Robineau-Desvoidy, Aldrichina Townsend, Hemipyrellia Townsend, Achoetandrus Bezzi, Protophormia Townsend, Chrysomya Robineau-Desvoidy, Lucilia Robineau-Desvoidy, Helicophagella Enderlein, and Boettcherisca Rohdendorf), and 12 species [Musca domestica (Linnaeus), Hydrotaea (Ophyra) capensis (Wiedemann), Lucilia caesar (Linnaeus), Lucilia illustris (Meigen), Aldrichina graham (Aldrich), Hemipyrellia ligurriens, Achoetandrus (Chrysomya) rufifacies (Macquary), Protophormia terraenovae (Robineau-Desvoidy), Chrysomya megacephala (Fabricius), Lucilia sericata (Meigen), Helicophagella melanura (Meigen), and Boettcherisca peregrine (Robineau-Desvoidy)]. The genus of the sarcosaphagous flies can be identified by 278 bp gene sequence analysis of CO I in mtDNA. This method is rapid, convenient and precise.

Molecular Epidemiology of Japanese Encephalitis Virus, Taiwan

Molecular Epidemiology of Japanese Encephalitis Virus, Taiwan