Enhanced precision in the analysis of randomized trials with ordinal outcomes

We present a general method for estimating the effect of a treatment on an ordinal outcome in randomized trials. The method is robust in that it does not rely on the proportional odds assumption. Our estimator leverages information in prognostic baseline variables, and has all of the following properties: (i) it is consistent; (ii) it is locally efficient; (iii) it is guaranteed to have equal or better asymptotic precision than both the inverse probability-weighted and the unadjusted estimators. To the best of our knowledge, this is the first estimator of the causal relation between a treatment and an ordinal outcome to satisfy these properties. We demonstrate the estimator in simulations based on resampling from a completed randomized clinical trial of a new treatment for stroke; we show potential gains of up to 39% in relative efficiency compared to the unadjusted estimator. The proposed estimator could be a useful tool for analyzing randomized trials with ordinal outcomes, since existing methods either rely on model assumptions that are untenable in many practical applications, or lack the efficiency properties of the proposed estimator. We provide R code implementing the estimator.

Economic analysis of microaerobic removal of H2S from biogas in full-scale sludge digesters

The application of microaerobic conditions during sludge digestion has been proven to be an efficient method for H2S removal from biogas. In this study, three microaerobic treatments were considered as an alternative to the technique of biogas desulfurization applied (FeCl3 dosing to the digesters) in a WWTP comprising three full-scale anaerobic reactors treating sewage sludge, depending on the reactant: pure O2 from cryogenic tanks, concentrated O2 from PSA generators, and air. These alternatives were compared in terms of net present value (NPV) with a fourth scenario consisting in the utilization of iron-sponge-bed filter inoculated with thiobacteria. The analysis revealed that the most profitable alternative to FeCl3 addition was the injection of concentrated O2 (0.0019 €/m(3) biogas), and this scenario presented the highest robustness towards variations in the price of FeCl3, electricity, and in the H2S concentration.

Evaluation of the Effect of a Continuous Treatment: A Machine Learning Approach with an Application to Treatment for Traumatic Brain Injury

For a continuous treatment, the generalised propensity score (GPS) is defined as the conditional density of the treatment, given covariates. GPS adjustment may be implemented by including it as a covariate in an outcome regression. Here, the unbiased estimation of the dose-response function assumes correct specification of both the GPS and the outcome-treatment relationship. This paper introduces a machine learning method, the 'Super Learner', to address model selection in this context. In the two-stage estimation approach proposed, the Super Learner selects a GPS and then a dose-response function conditional on the GPS, as the convex combination of candidate prediction algorithms. We compare this approach with parametric implementations of the GPS and to regression methods. We contrast the methods in the Risk Adjustment in Neurocritical care cohort study, in which we estimate the marginal effects of increasing transfer time from emergency departments to specialised neuroscience centres, for patients with acute traumatic brain injury. With parametric models for the outcome, we find that dose-response curves differ according to choice of specification. With the Super Learner approach to both regression and the GPS, we find that transfer time does not have a statistically significant marginal effect on the outcomes.

Targeted Maximum Likelihood Estimation using Exponential Families

Targeted maximum likelihood estimation (TMLE) is a general method for estimating parameters in semiparametric and nonparametric models. The key step in any TMLE implementation is constructing a sequence of least-favorable parametric models for the parameter of interest. This has been done for a variety of parameters arising in causal inference problems, by augmenting standard regression models with a "clever-covariate." That approach requires deriving such a covariate for each new type of problem; for some problems such a covariate does not exist. To address these issues, we give a general TMLE implementation based on exponential families. This approach does not require deriving a clever-covariate, and it can be used to implement TMLE for estimating any smooth parameter in the nonparametric model. A computational advantage is that each iteration of TMLE involves estimation of a parameter in an exponential family, which is a convex optimization problem for which software implementing reliable and computationally efficient methods exists. We illustrate the method in three estimation problems, involving the mean of an outcome missing at random, the parameter of a median regression model, and the causal effect of a continuous exposure, respectively. We conduct a simulation study comparing different choices for the parametric submodel. We find that the choice of submodel can have an important impact on the behavior of the estimator in finite samples.

A feasibility study on the bioconversion of CO2 and H2 to biomethane by gas sparging through polymeric membranes

In this study, the potential of a pilot hollow-fiber membrane bioreactor for the conversion of H2 and CO2 to CH4 was evaluated. The system transformed 95% of H2 and CO2 fed at a maximum loading rate of 40.2 [Formula: see text] and produced 0.22m(3) of CH4 per m(3) of H2 fed at thermophilic conditions. H2 mass transfer to the liquid phase was identified as the limiting step for the conversion, and kLa values of 430h(-1) were reached in the bioreactor by sparging gas through the membrane module. A simulation showed that the bioreactor could upgrade biogas at a rate of 25m(3)/mR(3)d, increasing the CH4 concentration from 60 to 95%v. This proof-of-concept study verified that gas sparging through a membrane module can efficiently transfer H2 from gas to liquid phase and that the conversion of H2 and CO2 to biomethane is feasible on a pilot scale at noteworthy load rates.

Variable importance and prediction methods for longitudinal problems with missing variables

We present prediction and variable importance (VIM) methods for longitudinal data sets containing continuous and binary exposures subject to missingness. We demonstrate the use of these methods for prognosis of medical outcomes of severe trauma patients, a field in which current medical practice involves rules of thumb and scoring methods that only use a few variables and ignore the dynamic and high-dimensional nature of trauma recovery. Well-principled prediction and VIM methods can provide a tool to make care decisions informed by the high-dimensional patient's physiological and clinical history. Our VIM parameters are analogous to slope coefficients in adjusted regressions, but are not dependent on a specific statistical model, nor require a certain functional form of the prediction regression to be estimated. In addition, they can be causally interpreted under causal and statistical assumptions as the expected outcome under time-specific clinical interventions, related to changes in the mean of the outcome if each individual experiences a specified change in the variable (keeping other variables in the model fixed). Better yet, the targeted MLE used is doubly robust and locally efficient. Because the proposed VIM does not constrain the prediction model fit, we use a very flexible ensemble learner (the SuperLearner), which returns a linear combination of a list of user-given algorithms. Not only is such a prediction algorithm intuitive appealing, it has theoretical justification as being asymptotically equivalent to the oracle selector. The results of the analysis show effects whose size and significance would have been not been found using a parametric approach (such as stepwise regression or LASSO). In addition, the procedure is even more compelling as the predictor on which it is based showed significant improvements in cross-validated fit, for instance area under the curve (AUC) for a receiver-operator curve (ROC). Thus, given that 1) our VIM applies to any model fitting procedure, 2) under assumptions has meaningful clinical (causal) interpretations and 3) has asymptotic (influence-curve) based robust inference, it provides a compelling alternative to existing methods for estimating variable importance in high-dimensional clinical (or other) data.

Discussion of Identification, Estimation and Approximation of Risk under Interventions that Depend on the Natural Value of Treatment Using Observational Data, by Jessica Young, Miguel Hernán, and James Robins

Young, Hernán, and Robins consider the mean outcome under a dynamic intervention that may rely on the natural value of treatment. They first identify this value with a statistical target parameter, and then show that this statistical target parameter can also be identified with a causal parameter which gives the mean outcome under a stochastic intervention. The authors then describe estimation strategies for these quantities. Here we augment the authors' insightful discussion by sharing our experiences in situations where two causal questions lead to the same statistical estimand, or the newer problem that arises in the study of data adaptive parameters, where two statistical estimands can lead to the same estimation problem. Given a statistical estimation problem, we encourage others to always use a robust estimation framework where the data generating distribution truly belongs to the statistical model. We close with a discussion of a framework which has these properties.

Estimating population treatment effects from a survey subsample

We considered the problem of estimating an average treatment effect for a target population using a survey subsample. Our motivation was to generalize a treatment effect that was estimated in a subsample of the National Comorbidity Survey Replication Adolescent Supplement (2001-2004) to the population of US adolescents. To address this problem, we evaluated easy-to-implement methods that account for both nonrandom treatment assignment and a nonrandom 2-stage selection mechanism. We compared the performance of a Horvitz-Thompson estimator using inverse probability weighting and 2 doubly robust estimators in a variety of scenarios. We demonstrated that the 2 doubly robust estimators generally outperformed inverse probability weighting in terms of mean-squared error even under misspecification of one of the treatment, selection, or outcome models. Moreover, the doubly robust estimators are easy to implement and provide an attractive alternative to inverse probability weighting for applied epidemiologic researchers. We demonstrated how to apply these estimators to our motivating example.

Hybrid periodic mesoporous organosilica designed to improve the properties of immobilized enzymes

Hybrid organosilica supports synthesized with pore size adjusted to enzyme dimensions provide high stability in organic solvent systems and prevent leaching.

Sensitivity analysis for causal inference under unmeasured confounding and measurement error problems

In this article, we present a sensitivity analysis for drawing inferences about parameters that are not estimable from observed data without additional assumptions. We present the methodology using two different examples: a causal parameter that is not identifiable due to violations of the randomization assumption, and a parameter that is not estimable in the nonparametric model due to measurement error. Existing methods for tackling these problems assume a parametric model for the type of violation to the identifiability assumption and require the development of new estimators and inference for every new model. The method we present can be used in conjunction with any existing asymptotically linear estimator of an observed data parameter that approximates the unidentifiable full data parameter and does not require the study of additional models.

Assessing the causal effect of policies: an example using stochastic interventions

Assessing the causal effect of an exposure often involves the definition of counterfactual outcomes in a hypothetical world in which the stochastic nature of the exposure is modified. Although stochastic interventions are a powerful tool to measure the causal effect of a realistic intervention that intends to alter the population distribution of an exposure, their importance to answer questions about plausible policy interventions has been obscured by the generalized use of deterministic interventions. In this article, we follow the approach described in Díaz and van der Laan (2012) to define and estimate the effect of an intervention that is expected to cause a truncation in the population distribution of the exposure. The observed data parameter that identifies the causal parameter of interest is established, as well as its efficient influence function under the non-parametric model. Inverse probability of treatment weighted (IPTW), augmented IPTW and targeted minimum loss-based estimators (TMLE) are proposed, their consistency and efficiency properties are determined. An extension to longitudinal data structures is presented and its use is demonstrated with a real data example.

Transcriptional analysis of intramuscular fatty acid composition in the longissimus thoracis muscle of Iberian × Landrace back-crossed pigs

This study aimed at identifying differential gene expression conditional on the fatty acid profile of the longissimus thoracis (Lt) muscle, a prime cut of economic relevance for fresh and cured pork production. A population of 110 Iberian (25%) × Landrace (75%) back-crossed pigs was used, because these two breeds exhibit extreme profiles of intramuscular saturated fatty acid, monounsaturated fatty acid (MUFA) and polyunsaturated fatty acid (PUFA) contents. Total RNA from Lt muscle was individually hybridized to GeneChip Porcine Genome arrays (Affymetrix). A principal component analysis was performed with data from the 110 animals to select 40 extreme animals based on the total fatty acid profile and the MUFA composition (MAP). Comparison of global transcription levels between extreme fatty acid profile pigs (n = 40) resulted in 219 differentially expressed probes (false discovery rate <0.10). Gene ontology, pathway and network analysis indicated that animals with higher percentages of PUFA exhibit a shift toward a more oxidative muscular metabolism state, with a raise in mitochondria function (PPARGC1A, ATF2), fatty acid uptake and oxidation (FABP5, MGLL). On the other hand, 87 probes were differentially expressed between MUFA composition groups (n = 40; false discovery rate <0.10). In particular, muscles rich in n-7 MUFA expressed higher levels of genes involved in lipid metabolism (GLUL, CRAT, PLA2G15) and lower levels of fatty acid elongation genes (ELOVL5). Moreover, the chromosomal position of FABP5, PAQR3, MGLL, PPARGC1A, GLUL and ELOVL5 co-localized with very relevant QTL for fat deposition and composition described in the same resource population. This study represents a complementary approach to identifying genes underlying these QTL effects.

First Report of Crown Rot of Strawberry Caused by Macrophomina phaseolina in Chile

In recent years, an increase of collapsed and dead strawberry plants has been observed in several fields in central Chile, specifically in San Pedro, Melipilla, an important area for strawberry cultivation in the country. To determine the causal agent of the disease and the extent of the problem, 25 sample sites of 1 ha each, distributed in different San Pedro zones, were surveyed at the end of the 2011 season (from December 2010 to February 2011). Cross sections of the crowns of symptomatic strawberry plants showed necrotic tissue and brown-red to dark brown areas on the vascular ring. Samples of the affected crowns were superficially disinfested and plated on potato dextrose agar with 200 μg/ml of streptomycin sulfate. Dark gray colonies were observed after 7 days of incubation at 24°C. Pure cultures of the pathogen showed aerial mycelium and abundant dark oblong sclerotia. Fifty sclerotia were measured, averaging 120 × 74 μm. Twenty-one isolates were identified molecularly utilizing the species specific primers MpKFI and MpKRI (2) that yielded a 350-bp fragment. The amplified DNA fragments were sequenced and BLAST analysis showed a 99% nucleotide sequence identity with Macrophomina phaseolina (GeneBank Accession No JX535007.1). Both morphological and molecular analyses confirmed that the isolated species corresponded to M. phaseolina, causal agent of crown and root rot in strawberry. Four representative isolates were selected to conduct pathogenicity tests. Inoculum was prepared by incubating the pathogen for 28 days at 20°C in sterilized oat seeds. Pots of 1.5 liters were filled with a mixed substrate of peat and perlite (2:1), amended with inoculated oats at 9 g per liter of substrate. 'Camarosa' strawberry plants were planted and grown in a glasshouse for 1 month. Six replicated plants per isolate and six plants growing on non-inoculated substrate were left as controls. Ninety-five percent of the inoculated plants showed wilt and collapse symptoms 22 days after transplant, whereas no symptoms were observed in the control plants. M. phaseolina was reisolated from the crowns of symptomatic plants, fulfilling Koch's postulates. The pathogen was isolated from symptomatic strawberry plants in 14 of the 25 sampled sites. Although M. phaseolina was described previously on other crops in Chile, to our knowledge, this is the first report of M. phaseolina causing crown rot of strawberry. The disease has been recently reported in Spain, the United States, and Argentina (1,3,4). References: (1) M. Avilés et al. Plant Pathol. 57:382, 2008. (2) B. Babu et al. Mycologia 99:797, 2007. (3) O. Baino et al. Plant Dis. 95:1477, 2011. (4) S. Koike. Plant Dis. 92:1253, 2008.

Comparison of different vaccination schedules for sustaining the immune response against porcine reproductive and respiratory syndrome virus

In order to better understand how immunization against porcine reproductive and respiratory syndrome virus (PRRSV) can be improved using commercial vaccines, different strategies of immunization were applied in the field using an inactivated vaccine (INV), a modified live vaccine (MLV) or a combination of the two and the responses compared. In experiment 1 (E1), 21 piglets were distributed in three groups. Group A was vaccinated with a commercial INV at 2.5, 3.5 and 6.5 months old; group B pigs received the INV at 1.5, 2.5, 5.5 and 6.5 months old, while pigs in group C were kept as unvaccinated controls. At 7.5 months of age all pigs were challenged with PRRSV and followed for 21 days. In experiment 2 (E2), 32 piglets were distributed evenly in four groups. Groups A, B and C were vaccinated with a commercial MLV at 1.5 months old, while group D pigs were kept as controls. At 4.5 months old, groups A and C received the INV while B received a second MLV, 1 month later group C pigs received a third INV. At 6.5 months old all pigs were challenged as in E1. In both experiments, total antibodies, neutralizing antibodies (NA) and cell-mediated immunity (CMI) were evaluated, and viraemia was determined after challenge. In E1, immunization with an INV induced high interferon-γ responses after the second and subsequent vaccinations. Development of NA after challenge was faster in INV vaccinated pigs compared to unvaccinated controls. In E2, re-vaccination with INV induced NA responses similar to re-vaccination with MLV; however, a significant increase in NA titres after challenge was only detected in group C pigs. The use of combined protocols (MLV+INV) was superior to the use of MLV alone in inducing cell mediated immunity. In conclusion, the highest immune responses against PRRSV after a single shot were achieved with MLV; after that, INV re-vaccination should be considered as the best strategy to induce significant boosters.

Association analysis with lipid traits of 2 candidate genes (LRP12 and TRIB1) mapping to a SSC4 QTL for serum triglyceride concentration in pigs

The performance of a genome scan for serum lipid traits at 45 and 190 d in 5 half-sib families of Duroc pigs allowed us to detect several pig chromosomal regions with significant effects on these phenotypes. In the current work, we aimed to refine the position of 1 chromosome 4 (SSC4) genome-wide significant QTL for serum triglyceride concentration at 190 d. Genotyping of 4 additional microsatellites allowed reduction of the 90% confidence interval of this QTL to the genomic interval between markers SW2409 and SW839. Sequencing experiments were performed to characterize the variability of 2 lipid-related genes, the lipoprotein receptor-related protein 12 (LRP12) and tribbles homolog 1 (TRIB1) loci, that map to this region. In this way, 2 (c.771A > G and c.1101A > G) and 1 (c.*156_157del) polymorphisms were identified at the LRP12 coding region and TRIB1 3' untranslated region, respectively. Association analyses between LRP12 and TRIB1 genotypes did not reveal any significant effect on serum lipid concentrations, suggesting that variation of these two loci does not explain the segregation of the SSC4 QTL. However, highly significant associations were observed for gluteus medius saturated fatty acid content (LRP12 c.1101A > G, P = 0.0006; TRIB1 c.*156_157del, P = 0.0003). In the light of these and other findings, the potential involvement of LRP12 and TRIB1 in muscle lipid metabolism deserves to be further explored.

Quality pork genes and meat production

Functional genomics, including analysis of the transcriptome and proteome, provides new opportunities for understanding the molecular processes in muscle and how these influence its conversion to meat. The Quality Pork Genes project was established to identify genes associated with variation in different aspects of raw material (muscle) quality and to then develop genetic tools that could be utilized to improve this quality. DNA polymorphisms identified in the porcine PRKAG3 and CAST genes illustrate the impact that such tools can have in improving meat quality. The resources developed in Quality Pork Genes provide the basis for identifying more of these tools.