BackgroundClinical researchers have often preferred to use a fixed effects model for the primary interpretation of a meta-analysis. Heterogeneity is usually assessed via the well known Q and I2 statistics, along with the random effects estimate they imply. In recent years, alternative methods for quantifying heterogeneity have been proposed, that are based on a 'generalised' Q statistic.MethodsWe review 18 IPD meta-analyses of RCTs into treatments for cancer, in order to quantify the amount of heterogeneity present and also to discuss practical methods for explaining heterogeneity.ResultsDiffering results were obtained when the standard Q and I2 statistics were used to test for the presence of heterogeneity. The two meta-analyses with the largest amount of heterogeneity were investigated further, and on inspection the straightforward application of a random effects model was not deemed appropriate. Compared to the standard Q statistic, the generalised Q statistic provided a more accurate platform for estimating the amount of heterogeneity in the 18 meta-analyses.ConclusionsExplaining heterogeneity via the pre-specification of trial subgroups, graphical diagnostic tools and sensitivity analyses produced a more desirable outcome than an automatic application of the random effects model. Generalised Q statistic methods for quantifying and adjusting for heterogeneity should be incorporated as standard into statistical software. Software is provided to help achieve this aim.
Nolwenn Callac, Carolane Giraud, Viviane Boulo
et al.
Background Aquacultured animals are reared in water hosting various microorganisms with which they are in close relationships during their whole lifecycle as some of these microorganisms can be involved in their host’s health or physiology. In aquaculture hatcheries, understanding the interactions existing between the natural seawater microbiota, the rearing water microbiota, the larval stage and the larval health status, may allow the establishment of microbial proxies to monitor the rearing ecosystems. Indeed, these proxies could help to define the optimal microbiota for shrimp larval development and could ultimately help microbial management. Methods In this context, we monitored the daily composition of the active microbiota of the rearing water in a hatchery of the Pacific blue shrimp Penaeus stylirostris. Two distinct rearing conditions were analyzed; one with antibiotics added to the rearing water and one without antibiotics. During this rearing, healthy larvae with a high survival rate and unhealthy larvae with a high mortality rate were observed. Using HiSeq sequencing of the V4 region of the 16S rRNA gene of the water microbiota, coupled with zootechnical and statistical analysis, we aimed to distinguish the microbial taxa related to high mortality rates at a given larval stage. Results We highlight that the active microbiota of the rearing water is highly dynamic whatever the larval survival rate. A clear distinction of the microbial composition is shown between the water harboring heathy larvae reared with antibiotics versus the unhealthy larvae reared without antibiotics. However, it is hard to untangle the effects of the antibiotic addition and of the larval death on the active microbiota of the rearing water. Various active taxa of the rearing water are specific to a given larval stage and survival rate except for the zoea with a good survival rate. Comparing these communities to those of the lagoon, it appears that many taxa were originally detected in the natural seawater. This highlights the great importance of the microbial composition of the lagoon on the rearing water microbiota. Considering the larval stage and larval survival we highlight that several genera: Nautella, Leisingera, Ruegerira , Alconivorax , Marinobacter and Tenacibaculum , could be beneficial for the larval survival and may, in the rearing water, overcome the r-strategist microorganisms and/or putative pathogens. Members of these genera might also act as probiotics for the larvae. Marivita , Aestuariicocccus, HIMB11 and Nioella , appeared to be unfavorable for the larval survival and could be associated with upcoming and occurring larval mortalities. All these specific biomarkers of healthy or unhealthy larvae, could be used as early routine detection proxies in the natural seawater and then during the first days of larval rearing, and might help to manage the rearing water microbiota and to select beneficial microorganisms for the larvae.
Surveillance of viral pathogens in both point-of-care and clinical settings is imperative to preventing the widespread propagation of disease—undetected viral outbreaks can pose dire health risks on a large scale. Thus, portable, accessible, and reliable biosensors are necessary for proactive measures. Polymeric microparticles have recently gained popularity for their size, surface area, and versatility, which make them ideal biosensing tools. This review cataloged recent investigations on polymeric microparticle-based detection platforms across eight virus families. These microparticles were used as labels for detection (often with fluorescent microparticles) and for capturing viruses for isolation or purification (often with magnetic microparticles). We also categorized all methods by the characteristics, materials, conjugated receptors, and size of microparticles. Current approaches were compared, addressing strengths and weaknesses in the context of virus detection. In-depth analyses were conducted for each virus family, categorizing whether the polymeric microparticles were used as labels, for capturing, or both. We also summarized the types of receptors conjugated to polymeric microparticles for each virus family.
This paper aims to provide an introduction to a basic form of the Q-tensor approach to modelling liquid crystals, which has seen increased interest in recent years. The increase in interest in this type of modelling approach has been driven by investigations into the fundamental nature of defects and new applications of liquid crystals such as bistable displays and colloidal systems for which a description of defects and disorder is essential. The work in this paper is not new research, rather it is an introductory guide for anyone wishing to model a system using such a theory. A more complete mathematical description of this theory, including a description of flow effects, can be found in numerous sources but the books by Virga and Sonnet and Virga are recommended. More information can be obtained from the plethora of papers using such approaches, although a general introduction for the novice is lacking. The first few sections of this paper will detail the development of the Q-tensor approach for nematic liquid crystalline systems and construct the free energy and governing equations for the mesoscopic dependent variables. A number of device surface treatments are considered and theoretical boundary conditions are specified for each instance. Finally, an example of a real device is demonstrated.
We consider model-free reinforcement learning for infinite-horizon discounted Markov Decision Processes (MDPs) with a continuous state space and unknown transition kernel, when only a single sample path under an arbitrary policy of the system is available. We consider the Nearest Neighbor Q-Learning (NNQL) algorithm to learn the optimal Q function using nearest neighbor regression method. As the main contribution, we provide tight finite sample analysis of the convergence rate. In particular, for MDPs with a $d$-dimensional state space and the discounted factor $\gamma \in (0,1)$, given an arbitrary sample path with "covering time" $ L $, we establish that the algorithm is guaranteed to output an $\varepsilon$-accurate estimate of the optimal Q-function using $\tilde{O}\big(L/(\varepsilon^3(1-\gamma)^7)\big)$ samples. For instance, for a well-behaved MDP, the covering time of the sample path under the purely random policy scales as $ \tilde{O}\big(1/\varepsilon^d\big),$ so the sample complexity scales as $\tilde{O}\big(1/\varepsilon^{d+3}\big).$ Indeed, we establish a lower bound that argues that the dependence of $ \tilde{\Omega}\big(1/\varepsilon^{d+2}\big)$ is necessary.