Searching for convergence in phylogenetic Markov chain Monte Carlo

Markov chain Monte Carlo (MCMC) is a methodology that is gaining widespread use in the phylogenetics community and is central to phylogenetic software packages such as MrBayes. An important issue for users of MCMC methods is how to select appropriate values for adjustable parameters such as the length of the Markov chain or chains, the sampling density, the proposal mechanism, and, if Metropolis-coupled MCMC is being used, the number of heated chains and their temperatures. Although some parameter settings have been examined in detail in the literature, others are frequently chosen with more regard to computational time or personal experience with other data sets. Such choices may lead to inadequate sampling of tree space or an inefficient use of computational resources. We performed a detailed study of convergence and mixing for 70 randomly selected, putatively orthologous protein sets with different sizes and taxonomic compositions. Replicated runs from multiple random starting points permit a more rigorous assessment of convergence, and we developed two novel statistics, delta and epsilon, for this purpose. Although likelihood values invariably stabilized quickly, adequate sampling of the posterior distribution of tree topologies took considerably longer. Our results suggest that multimodality is common for data sets with 30 or more taxa and that this results in slow convergence and mixing. However, we also found that the pragmatic approach of combining data from several short, replicated runs into a metachain to estimate bipartition posterior probabilities provided good approximations, and that such estimates were no worse in approximating a reference posterior distribution than those obtained using a single long run of the same length as the metachain. Precision appears to be best when heated Markov chains have low temperatures, whereas chains with high temperatures appear to sample trees with high posterior probabilities only rarely. [Bayesian phylogenetic inference; heating parameter; Markov chain Monte Carlo; replicated chains.]

Biodiversity conservation planning tools: Present status and challenges for the future

Species extinctions and the deterioration of other biodiversity features worldwide have led to the adoption of systematic conservation planning in many regions of the world. As a consequence, various software tools for conservation planning have been developed over the past twenty years. These, tools implement algorithms designed to identify conservation area networks for the representation and persistence of biodiversity features. Budgetary, ethical, and other sociopolitical constraints dictate that the prioritized sites represent biodiversity with minimum impact on human interests. Planning tools are typically also used to satisfy these criteria. This chapter reviews both the concepts and technical choices that underlie the development of these tools. Conservation planning problems can be formulated as optimization problems, and we evaluate the suitability of different algorithms for their solution. Finally, we also review some key issues associated with the use of these tools, such as computational efficiency, the effectiveness of taxa and abiotic parameters at choosing surrogates for biodiversity, the process of setting explicit targets of representation for biodiversity surrogates, and

Postharvest characteristics and handling of litchi fruit - an overview

Litchi ( Litchi chinensis Sonn.) is a tropical to subtropical crop that originated in South-East Asia. Litchi fruit are prized on the world market for their flavour, semi-translucent white aril and attractive red skin. Litchi is now grown commercially in many countries and production in Australia, China, Israel, South Africa and Thailand has expanded markedly in recent years. Increased production has made significant contributions to economic development in these countries, especially those in South-East Asia. Non-climacteric litchi fruit are harvested at their visual and organoleptic optimum. They are highly perishable and, consequently, have a short life that limits marketability and potential expansion of demand. Pericarp browning and pathological decay are common and important defects of harvested litchi fruit. Postharvest technologies have been developed to reduce these defects. These technologies involve cooling and heating the fruit, use of various packages and packaging materials and the application of fungicides and other chemicals. Through the use of fungicides and refrigeration, litchi fruit have a storage life of about 30 days. However, when they are removed from storage, their shelf life at ambient temperature is very short due to pericarp browning and fruit rotting. Low temperature acclimation or use of chitsoan as a coating can extend the shelf life. Sulfur dioxide fumigation effectively reduces pericarp browning, but approval from Europe, Australia and Japan for this chemical is likely to be withdrawn due to concerns over sulfur residues in fumigated fruit. Thus, sulfur-free postharvest treatments that maintain fruit skin colour are increasingly important. Alternatives to SO2 fumigation for control of pericarp browning and fruit rotting are pre-storage pathogen management, anoxia treatment, and dipping in 2% hydrogen chloride solution for 6-8 min following storage at 0 degrees C. Insect disinfestation has become increasingly important for the expansion of export markets because of quarantine issues associated with some fruit fly species. Thus, effective disinfestation protocols need to be developed. Heat treatment has shown promise as a quarantine technology, but it injures pericarp tissue and results in skin browning. However, heat treatment can be combined with an acid dip treatment that inhibits browning. Therefore, the primary aim of postharvest litchi research remains the achievement of highly coloured fruit which is free of pests and disease. Future research should focus on disease control before harvest, combined acid and heat treatments after harvest and careful temperature management during storage and transport.

Isolates of 'Candidatus Nostocoida limicola' Blackall et al. 2000 should be described as three novel species of the genus Tetrasphaera, as Tetrasphaera jenkinsii sp nov., Tetrasphaera vanveenii sp nov and Tetrasphaera veronensis sp nov.

Despite differences in their morphologies, comparative analyses of 16S rRNA gene sequences revealed high levels of similarity (> 94 %) between strains of the filamentous bacterium 'Candidatus Nostocoida limicola' and the cocci Tetrasphaera australiensis and Tetrasphaera japonica and the rod Tetrasphaera elongata, all isolated from activated sludge. These sequence data and their chemotaxonomic characters, including cell wall, menaquinone and lipid compositions and fingerprints of their 16S-23S rRNA intergenic regions, support the proposition that these isolates should be combined into a single genus containing six species, in the family Intrasporangiaceae in the Actinobacteria. This suggestion receives additional support from DNA-DNA hybridization data and when partial sequences of the rpoC1 gene are compared between these strains. Even though few phenotypic characterization data were obtained for these slowly growing isolates, it is proposed, on the basis of the extensive chemotaxonomic and molecular evidence presented here, that 'Candidatus N. limicola' strains Ben 17, Ben 18, Ben 67, Ben 68 and Ben 74 all be placed into the species Tetrasphaera jenkinsii sp. nov. (type strain Ben 74(T) = DSM 17519(T) = NCIMB 14128(T)), 'Candidatus N. limicola' strain Ben 70 into Tetrasphaera vanveenii sp. nov. (type strain Ben 70(T) = DSM 17518(T) = NCIMB 14127(T)) and 'Candidatus N. limicola' strains Ver 1 and Ver 2 into Tetrasphaera veronensis sp. nov. (type strain Ver 1(T) = DSM 17520(T) = NCIMB 14129(T)).

Health system reform in Mexico 5: Assessing the effect of the 2001-06 Mexican health reform: an interim report card

Since 2001, Mexico has been designing, legislating, and implementing a major health-system reform. A key component was the creation of Seguro Popular, which is intended to expand insurance coverage over 7 years to uninsured people, nearly half the total population at the start of 2001. The reform included five actions: legislation of entitlement per family affiliated which, with full implementation, will increase public spending on health by 0.8-1.0% of gross domestic product; creation of explicit benefits packages; allocation of monies to decentralised state ministries of health in proportion to number of families affiliated; division of federal resources flowing to states into separate funds for personal and non-personal health services; and creation of a fund to protect families against catastrophic health expenditures. Using the WHO health-systems framework, we used a wide range of datasets to assess the effect of this reform on different dimensions of the health system. Key findings include: affiliation is preferentially reaching the poor and the marginalised communities; federal non-social security expenditure in real per-head terms increased by 38% from 2000 to 2005; equity of public-health expenditure across states improved; Seguro Popular affiliates used more inpatient and outpatient services than uninsured people; effective coverage of 11 interventions has improved between 2000 and 2005-06; inequalities in effective coverage across states and wealth deciles has decreased over this period; catastrophic expenditures for Seguro Popular affiliates are lower than for uninsured people even though use of services has increased. We present some lessons for Mexico based on this interim evaluation and explore implications for other countries considering health reforms.

Parthenium weed: a potential major weed for agro-ecosystems in Pakistan

Parthenium weed (Parthenium hysterophorus L.) is a new and potentially major weed in Pakistan. This weed, originating from central America, is now a major weed in many regions of the world including Eastern Africa, India, parts of South East Asia and Australia. Presumably its recent arrival in Pakistan has been due to its movement from India, but this has yet to be established. In Australia it has been present for about 50 years, in which time it has spread from isolated infestations to establish core populations in central Queensland with scattered and isolated plants occurring south into New South Wales and north-west into the Northern Territory. Its spread in Pakistan is likely to be much more rapid, but lessons learnt in Australia will be of great value for weed managers in Pakistan. This annual herb has the potential to spread to all medium rainfall rangeland, dairy and summer cropping areas in Pakistan. In Australia its main effect is upon livestock production, but it is also causing health concerns in regional communities. However, in India it has also had a significant impact in cropping systems. To help coordinate actions on its management in Australia, a National Weeds Program has created a Parthenium Weed Management Group (PWMG) and under this group a Parthenium Weed Research Group (PWRG) has been formed. Funding coming from this national program and other sources has supported the PWRG to undertake a collaborative and technology exchange research program in two main areas: 1) biology and ecology and 2) management; while the PWMG has focused on community awareness and the production of various extension and management packages. Research in the area of biology and ecology has included studies on the evaluation of competitive plants to displace parthenium weed, the use of process-based simulation models to monitor and predict future spread and abundance under present and future climate conditions, the effect of the weed on human health and the ecology of its seed bank. Management research has focussed on the development of biological control approaches using plant-feeding insects and pathogens. The effectiveness of biological control is also being monitored through long term studies on seed bank size and dynamics. The use of fire as another potential management tool is also being evaluated. In addition to this important research, an effort has also been made to spread the most important findings and management outcomes to the wider community through an extension and education program driven by the PWMG. These developments within Australia, in parthenium weed management, will be of great help to P

Development of a web-service-agents-based family wealth management system

The roiling financial markets, constantly changing tax law and increasing complexity of planning transaction increase the demand of aggregated family wealth management (FWM) services. However, current trend of developing such advisory systems is mainly focusing on financial or investment side. In addition, these existing systems lack of flexibility and are hard to be integrated with other organizational information systems, such as CRM systems. In this paper, a novel architecture of Web-service-agents-based FWM systems has been proposed. Multiple intelligent agents are wrapped as Web services and can communicate with each other via Web service protocols. On the one hand, these agents can collaborate with each other and provide comprehensive FWM advices. On the other hand, each service can work independently to achieve its own tasks. A prototype system for supporting financial advice is also presented to demonstrate the advances of the proposed Webservice- agents-based FWM system architecture.

Reflective web interface agent

Web interface agent is used with web browsers to assist users in searching and interactions with the WWW. It is used for a variety of purposes, such as web-enabled remote control, web interactive visualization, and e-commerce activities. User may be aware or unaware of its existence. The intelligence of interface agent consists in its capability of learning and decision-making in performing interactive functions on behalf of a user. However, since web is an open system environment, the reasoning mechanism in an agent should be able to adapt changes and make decisions on exceptional situations, and therefore use meta knowledge. This paper proposes a framework of Reflective Web Interface Agent (RWIA) that is to provide causal connections between the application interfaces and the knowledge model of the interface agent. A prototype is also implemented for the purpose of demonstration.

Boosting the tree augmented naive bayes classifier

The Tree Augmented Naïve Bayes (TAN) classifier relaxes the sweeping independence assumptions of the Naïve Bayes approach by taking account of conditional probabilities. It does this in a limited sense, by incorporating the conditional probability of each attribute given the class and (at most) one other attribute. The method of boosting has previously proven very effective in improving the performance of Naïve Bayes classifiers and in this paper, we investigate its effectiveness on application to the TAN classifier.

Improving support vector solutions by selecting a sequence of training subsets

In this paper we demonstrate that it is possible to gradually improve the performance of support vector machine (SVM) classifiers by using a genetic algorithm to select a sequence of training subsets from the available data. Performance improvement is possible because the SVM solution generally lies some distance away from the Bayes optimal in the space of learning parameters. We illustrate performance improvements on a number of benchmark data sets.