Building composite indices in the age of big data - Application to honey bee exposure to infectious and parasitic agents.

Pollinator insects play a crucial role in maintaining biodiversity and agricultural production worldwide. Yet they are subject to various infectious and parasitic agents (IPAs). To better assess their exposure to IPAs, discriminative and quantitative molecular methods have been developed. These tools produce large datasets that need to be summarised so as to be interpreted. In this paper, we described the calculation of three types of composite indices (numerical, ordinal, nominal) to characterize the honey bee exposure to IPAs in 128 European sites. Our summarizing methods are based on component-based factorial analyses. The indices summarised the dataset of eight IPAs quantified at two sampling times, into synthetic values providing different yet complementary information. Because our dataset included two sampling times, we used Multiple Factor Analysis (MFA) to synthetize the information. More precisely, the numerical and ordinal indices were generated from the first component of MFA, whereas the nominal index used the first main components of MFA combined with a clustering analysis (Hierarchical Clustering on components). The numerical index was easy to calculate and to be used in further statistical analyses. However, it contained only about 20% of the original information. Containing the same amount of original information, the ordinal index was much easier to interpret. These two indices summarised information in a unidimensional manner. Instead, the nominal index summarised information in a multidimensional manner, which retained much more information (94%). In the practical example, the three indices showed an antagonistic relationship between N. ceranae and DWV-B. These indices represented a toolbox where scientists could pick one composite index according to the aim pursued. Indices could be used in further statistical analyses but could also be used by policy makers and public instances to characterize a given sanitary situation at a site level for instance.

Investigation on eggshell apex abnormality (EAA) syndrome in France: isolation of Mycoplasma synoviae is frequently associated with Mycoplasma pullorum.

BACKGROUND: Mycoplasma synoviae (MS) is known to cause Eggshell Apex Abnormality (EAA) syndrome characterized by an altered shell surface with increased translucency on the apex. However, no large-scale studies have been conducted to obtain prevalence data of EAA and MS isolates associated to this syndrome. This manuscript reports the results of two field studies performed in the French poultry industry (2015-2017): focusing mainly on investigation of presence and prevalence of EAA in different types of laying hen flocks (phase 1), and isolation of MS strains from EAA-infected flocks (phase 2). RESULTS: The first survey included 77 farms of commercial layers in three French egg-production regions, hosting 40 flocks in alternative systems (ALT) and 56 in furnished cages (FC). Seven flocks (4 FC and 3 ALT) presented EAA clinical signs, giving a prevalence of 7.3% in this studied sample. A second independent field study was conducted to identify MS by in vitro cultivation and PCR in samples from 28 flocks with clinical signs of EAA. Different types of biological specimens were collected in EAA-affected flocks and submitted to the laboratory. M. synoviae was detected in 25/28 flocks, from both production systems (5/5 ALT and 20/23 FC). Detection of MS was significantly higher in tracheal swabs (59%) than in cloacal (10.5%), albumen (3.6%) and egg yolk (1.1%) swabs. It is worth to mention that attempts to clone MS from positive samples were often hampered by the presence of another Mycoplasma species, which showed fast growing behaviour in the selective media used in this study (Frey Medium 4 and Frey Medium 4 supplemented with erythromycin). The use of MALDI-TOF mass spectrometry in combination with next-generation sequencing (NGS) results allowed the identification of this fast growing mycoplasma as Mycoplasma pullorum, which was detected in 14 of the 25 (56%) MS-positive flocks. CONCLUSIONS: These results confirmed the presence of the EAA syndrome in MS-positive flocks of layers in France, reared in different regions and in different production systems (ALT and FC). Studies need to be conducted to test whether M. pullorum may influence the expression of clinical signs of EAA in MS-infected layer farms.

Application of multiblock modelling to identify key drivers for antimicrobial use in pig production in four European countries.

Antimicrobial use in pig farming is influenced by a range of risk factors, including herd characteristics, biosecurity level, farm performance, occurrence of clinical signs and vaccination scheme, as well as farmers' attitudes and habits towards antimicrobial use. So far, the effect of these risk factors has been explored separately. Using an innovative method called multiblock partial least-squares regression, this study aimed to investigate, in a sample of 207 farrow-to-finish farms from Belgium, France, Germany and Sweden, the relative importance of the six above mentioned categories or 'blocks' of risk factors for antimicrobial use in pig production. Four country separate models were developed; they showed that all six blocks provided useful contribution to explaining antimicrobial use in at least one country. The occurrence of clinical signs, especially of respiratory and nervous diseases in fatteners, was one of the largest contributing blocks in all four countries, whereas the effect of the other blocks differed between countries. In terms of risk management, it suggests that a holistic and country-specific mitigation strategy is likely to be more effective. However, further research is needed to validate our findings in larger and more representative samples, as well as in other countries.

Analysis of the influences on plumage condition in laying hens: How suitable is a whole body plumage score as an outcome?

An important indicator of the health and behavior of laying hens is their plumage condition. Various scoring systems are used, and various risk factors for feather damage have been described. Often, a summarized score of different body parts is used to describe the overall condition of the plumage of a bird. However, it has not yet been assessed whether such a whole body plumage score is a suitable outcome variable when analyzing the risk factors for plumage deterioration. Data collected within a German project on farms keeping laying hens in aviaries were analyzed to investigate whether and the extent to which information is lost when summarizing the scores of the separate body parts. Two models were fitted using multiblock redundancy analysis, in which the first model included the whole body score as one outcome variable, while the second model included the scores of the individual body parts as multiple outcome variables. Although basically similar influences could be discovered with both models, the investigation of the individual body parts allowed for consideration of the influences on each body part separately and for the identification of additional influences. Furthermore, ambivalent influences (a factor differently associated with 2 different outcomes) could be detected with this approach, and possible dilutive effects were avoided. We conclude that influences might be underestimated or even missed when modeling their explanatory power for an overall score only. Therefore, multivariate methods that allow for the consideration of individual body parts are an interesting option when investigating influences on plumage condition.

Escherichia coli Probiotic Strain ED1a in Pigs Has a Limited Impact on the Gut Carriage of Extended-Spectrum-ß-Lactamase-Producing E. coli.

Four trials were conducted to evaluate the impact of Escherichia coli probiotic strain ED1a administration to pigs on the gut carriage or survival in manure of extended-spectrum-ß-lactamase-producing E. coli Groups of pigs were orally inoculated with strain E. coli M63 carrying the blaCTX-M-1 gene (n = 84) or used as a control (n = 26). In the first two trials, 24 of 40 E. coli M63-inoculated pigs were given E. coli ED1a orally for 6 days starting 8 days after oral inoculation. In the third trial, 10 E. coli M63-inoculated pigs were given either E. coli ED1a or probiotic E. coli Nissle 1917 for 5 days. In the fourth trial, E. coli ED1a was given to a sow and its 12 piglets, and these 12 piglets plus 12 piglets that had not received E. coli ED1a were then inoculated with E. coli M63. Fecal shedding of cefotaxime-resistant Enterobacteriaceae (CTX-RE) was studied by culture, and blaCTX-M-1 genes were quantified by PCR. The persistence of CTX-RE in manure samples from inoculated pigs or manure samples inoculated in vitro with E. coli M63 with or without probiotics was studied. The results showed that E. coli M63 and ED1a were good gut colonizers. The reduction in the level of fecal excretion of CTX-RE in E. coli ED1a-treated pigs compared to that in nontreated pigs was usually less than 1 log10 CFU and was mainly observed during the probiotic administration period. The results obtained with E. coli Nissle 1917 did not differ significantly from those obtained with E. coli ED1a. CTX-RE survival did not differ significantly in manure samples with or without probiotic treatment. In conclusion, under our experimental conditions, E. coli ED1a and E. coli Nissle 1917 could not durably prevent CTX-RE colonization of the pig gut.

Virulence Genes in Expanded-Spectrum-Cephalosporin-Resistant and -Susceptible Escherichia coli Isolates from Treated and Untreated Chickens.

This study investigated antimicrobial resistance, screened for the presence of virulence genes involved in intestinal infections, and determined phylogenetic groups of Escherichia coli isolates from untreated poultry and poultry treated with ceftiofur, an expanded-spectrum cephalosporin. Results show that none of the 76 isolates appeared to be Shiga toxin-producing E. coli or enteropathogenic E. coli. All isolates were negative for the major virulence factors/toxins tested (ehxA, cdt, heat-stable enterotoxin [ST], and heat-labile enterotoxin [LT]). The few virulence genes harbored in isolates generally did not correlate with isolate antimicrobial resistance or treatment status. However, some of the virulence genes were significantly associated with certain phylogenetic groups.

Association of herd BHV-1 seroprevalence with respiratory disease in youngstock in Estonian dairy cattle.

The associations between herd bovine herpesvirus 1 (BHV-1) seroprevalence, along with other infectious and farm management factors with bovine respiratory disease (BRD) in dairy calves and heifers, were investigated. Serum samples from 103 dairy cattle herds were analyzed for antibodies against BHV-1, bovine respiratory syncytial virus (BRSV), bovine viral diarrhea virus (BVDV), and Mycoplasma bovis (M. bovis). A questionnaire was used to record herd management practices. A high occurrence of respiratory disease in unweaned calves was associated with low to moderate and high prevalence of BHV-1 among cows (OR=14.8, p=0.005 and OR=19.2, p=0.002, respectively) and positive BVDV status of a herd (OR=5.1, p=0.02). The presence of BVDV in a herd was related to a high incidence of respiratory disease in heifers 3-16 months old (OR=4.3, p=0.027). Based on the results of multiple correspondence analysis, holding youngstock separately from cows until pregnancy, introducing new animals and the activities of on-farm employees may contribute to a higher incidence of BRD.

Multiblock modelling to assess the overall risk factors for a composite outcome.

Research in epidemiology may be concerned with assessing risk factors for complex health issues described by several variables. Moreover, epidemiological data are usually organized in several blocks of variables, consisting of a block of variables to be explained and a large number of explanatory variables organized in meaningful blocks. Usual statistical procedures such as generalized linear models do not allow the explanation of a multivariate outcome, such as a complex disease described by several variables, with a single model. Moreover, it is not easy to take account of the organization of explanatory variables into blocks. Here we propose an innovative method in the multiblock modelling framework, called multiblock redundancy analysis, which is designed to handle most specificities of complex epidemiological data. Overall indices and graphical displays associated with different interpretation levels are proposed. The interest and relevance of multiblock redundancy analysis is illustrated using a dataset pertaining to veterinary epidemiology.

Risk factors for sanitary condemnation in broiler chickens and their relative impact: application of an original multiblock approach.

An innovative and well-adapted statistical method, called multiblock redundancy analysis, is proposed for a complex health-event analysis to account for the thematic block organization of variables. The outcome block contained the condemnation rates of 404 broiler chicken flocks, distinguishing infectious and traumatic condemnation categories. Explanatory variables were organized in blocks related to the different production stages (farm structure and routine husbandry practices; on-farm flock history and characteristics; catching, transport and lairage conditions; slaughterhouse and inspection features). The aim was to determine risk factors for both condemnation categories, and the relative impact of the different production stages on the whole condemnation rate. Results showed that significant factors were either specific to one condemnation category or related to both categories, and each of the explanatory blocks was involved in the explanation of infectious and traumatic condemnation rates. On-farm flock information explained 40% of the overall condemnation process whereas the other explanatory blocks had similar relative impacts.

Bayesian estimation of flock-level sensitivity of detection of Salmonella spp., Enteritidis and Typhimurium according to the sampling procedure in French laying-hen houses.

A study was carried out to estimate the prevalence of flocks infected by Salmonella spp., S. Enteritidis and S. Typhimurium in 521 French laying-hen farms from October 1st 2004 to September 30th 2005 as part of a European Union-wide baseline study to define targets for Salmonella reduction in member states. The sampling scheme prescribed and financed by the European Commission to detect Salmonella in laying-hen flocks was based on 2 dust-samples and 5 faeces-samples per farm. A latent-class Bayesian approach for correlated tests was used to estimate the sensitivity of detection of reduced sampling schemes corresponding to the 16 combinations of 2 dust- and 5 faeces-samples. For each model the full sampling scheme (7 samples) and the reduced protocol were considered as two correlated tests, the biological principle being identical and the reduced protocol being a subset of the full sampling scheme. As the observed apparent prevalence in cage flocks was higher than in other systems (barns, outdoor, or organic) these two sub-populations were considered separately. Bayesian estimation of posterior medians with 95% probability intervals for true prevalence in cage flocks were 0.34 (0.29; 0.39) and 0.13 (0.10; 0.18) for Salmonella spp. and Salmonella Enteritidis+Typhimurium respectively. In alternative flocks posterior medians with 95% probability intervals for true prevalence were 0.09 (0.06; 0.13) and 0.05 (0.03; 0.08) for Salmonella spp. and Salmonella Enteritidis+Typhimurium, respectively. In cage flocks Bayesian estimation of posterior distributions for sensitivity indicated that at least 5 samples, including 2 dust samples were necessary to attain comparable sensitivity levels to the full sampling scheme. In alternative flocks and for Salmonella spp. 6 samples were required to ensure a comparable sensitivity level to the full sampling scheme. Detection sensitivity was improved by increasing the number of dust samples in cage farms and by increasing the total number of samples whatever their type in alternative farms.

Validation of two commercial real-time RT-PCR kits for rapid and specific diagnosis of classical swine fever virus.

Two real-time RT-PCR kits, developed by LSI (TaqVet CSF) and ADIAGENE (Adiavet CSF), obtained an agreement to be commercialised in France, subject to conditions, defined by the French Classical Swine Fever (CSF) National Reference Laboratory. The producers were asked to introduce an internal control to check the RNA extraction efficacy. The different criteria assessed were sensitivity, "pestivirus specificity", reproducibility and ease of handling, using 189 different samples. These samples were either CSFV inactivated strains or blood/serum/organs collected from CSFV experimentally infected pigs or naturally infected wild boars. The reproducibility of the assays was confirmed by the analysis of a batch-to-batch panel control that was used for inter-laboratory tests involving nine laboratories. The two kits were also tested for the use in mass diagnostics and the results proved the kits to be suited using pools of blood, serum and tonsils. Moreover, a field evaluation, carried out on spleen samples collected from the CSF surveillance of wild boars in an area known to be infected and from domestic pigs at a slaughterhouse, confirmed the high sensitivity and specificity of the two kits. This step-by-step evaluation procedure confirmed that the two commercial CSF real-time RT-PCR kits have a higher predictive value than the current diagnostic standard, Virus Isolation.

Immune and protective abilities of ubiquitinated and non-ubiquitinated pseudorabies virus glycoproteins.

Plasmids encoding ubiquitinated (ubi-) or non-ubiquitinated (non-ubi-) glycoproteins of Pseudorabies virus (PRV) were used for vaccination of pigs. We found that the fusion of ubiquitin to viral glycoproteins increased their degradation in proteasomes in vitro, in which ubiquitin plays a key role. In the animals immunized with the plasmids encoding PRV ubi-glycoproteins and then challenged with PRV, we detected a slightly decreased cellular immune response on days 13 and 19 after immunization and a reduced nasal excretion of infectious virus on day 2 after the challenge. Afterwards, no effect of the ubiquitination of the glycoproteins on humoral or cellular immunity and on excretion of infectious virus was observed. Similarly, no effect of the ubiquitination on protective abilities of PRV glycoproteins was found.

Laboratory evaluation of a quantitative real-time reverse transcription PCR assay for the detection and identification of the four subgroups of avian metapneumovirus.

Avian metapneumovirus (AMPV) is an important pathogen causing respiratory diseases and egg drops in several avian species. Four AMPV subgroups have been identified. The laboratory diagnosis of AMPV infections relies on serological methods, on labour-intensive virus isolation procedures, and on recently developed subgroup specific reverse transcription PCR (RT-PCR) protocols. In the present study, both the specificity and sensitivity of a commercial real-time reverse transcription PCR (RRT-PCR) for the detection and identification of the four AMPV subgroups were evaluated. Fifteen non-AMPV avian viruses belonging to 7 genera and 32 AMPV belonging to the 4 subgroups were tested. No non-AMPV virus was detected, whereas all AMPV viruses were identified in agreement with their previous molecular and antigenic subgroup assignment. The sensitivity and quantitating ability of the RRT-PCR assay were determined using serial dilutions of RNA derived either from AMPV virus stocks or from runoff transcripts. In all cases, linear dose/responses were observed. The detection limits of the different subgroups ranged from 500 to 5000 RNA copies and from 0.03 to 3.16TCID50/ml. The results were reproducible under laboratory conditions, thus showing that quantitative RRT-PCR is a new and powerful tool for the rapid and sensitive detection, identification and quantitation of AMPVs.

Validation of a real-time RT-PCR assay for rapid and specific diagnosis of Classical Swine Fever virus.

Two new real-time RT-PCR kits developed by LSI (TaqVet CSF) and ADIAGENE (Adiavet CSF) obtained a manufacturing agreement in France during the past year. For that purpose, the Classical Swine Fever (CSF) National Reference Laboratory (NRL) planned a schedule of conditions to be fulfilled by commercial real-time RT-PCR assays. The producers were asked to introduce an internal control to check the RNA extraction efficacy. The different criteria assessed were: sensitivity, specificity, especially "pestivirus specificity", reproducibility and easy handling, using 187 different samples distributed in four different panels. These samples were either CSFV inactivated strains or organs collected from CSF experimental SPF infected pigs, or naturally infected wild boars. All these samples were previously tested for genome detection using an RT-nested PCR assay and for virus isolation on cell culture. The LSI TaqVet kit was used for the CSF surveillance of wild boars in an area known to be infected, during the winter of 2004-2005. This field evaluation was carried out on 4710 spleen samples. In summary, the new CSF real-time RT-PCR assays have a higher predictive value than the current diagnostic standard, Virus Isolation.

Multiplex PCR assay for detection of Streptococcus suis species and serotypes 2 and 1/2 in tonsils of live and dead pigs.

A PCR assay was developed for the detection of Streptococcus suis serotypes 2 and 1/2. This multiplex PCR is based on the amplification of the gene coding for 16S rRNA of S. suis and on the amplification of the cps2J gene coding for the capsule of S. suis serotypes 2 and 1/2. An internal control was constructed and added in this test to monitor the efficiency of amplification in each reaction. To evaluate the specificity of the test, 31 strains of other bacterial species related to S. suis or isolated from pigs and 42 strains of S. suis serotypes 1 and 3 to 34 were analyzed. The detection threshold of the test was 28 S. suis CFU/ml. The specificity and the sensitivity of the multiplex PCR test and the presence of an internal control allowed the analysis of biological samples without a culture step. The PCR assay was then applied to the detection of 14 S. suis serotype 1/2 strains, 88 S. suis serotype 2 strains isolated from pigs, and 25 S. suis serotype 2 strains isolated from humans. This test was also applied to analyze tonsil samples of pigs experimentally infected and carrier pigs without any symptoms.