基于抗体筛查的美国猪群流行性腹泻病毒真流行率分布推断

养殖家畜的冠状病毒,如猪流行性腹泻病毒(porcine epidemic diarrhea virus,PEDV),其基因序列与蝙蝠体内冠状病毒基因序列高度相似,可引起猪肠道感染并导致仔猪大量死亡,对动物健康危害极大.本研究基于对美国猪群PEDV流行毒株表观流行率的数据挖掘作为先验信息,以2019年美国生猪入境中国西南...

Bayesian Inference and Simulation on True Prevalence of Porcine Epidemic Diarrhea Virus in American Swine Population

The coronaviruses of livestock, such as porcine epidemic diarrhea virus-PEDV, have genetic sequences that are highly similar to those of coronaviruses in bats, which can cause intestinal infections in pigs and cause a large number of deaths of piglets, which is extremely harmful to the health of animals. This study is based on the data mining of the epidemiological prevalence of PEDV strains in the United States as a priori information, combined with antibody screening of US swine in 2019 at entry port as the knowledge update, Bayesian statistical inference were employed to estimate the true prevalence of PEDV in US swine populations. The results show that the median true prevalence (median) of PEDV in the US swine population is 0.005 22 (95% CI=0.000 424-0.022 5), and the prevalence distribution of the 95% upper confident limit (UCL) is 3%. The probability density statistical properties of the posterior distribution prevalence are highly right-skewed to infer that the true prevalence of PEDV in the U.S. pig herd is mainly concentrated between in the 2.5% percentile (P_2.5%=0.0002) and the median (P_50%=0.005 22), but due to the variability and difference among different pig herds in the United States, the PEDV true prevalence distribution value of a random sampling may have a small probability overflowing the distribution greater than 0.03 (P ≤ 2.5%). In addition, in terms of assay selection for port screening, it is recommended that N protein-based ELISA can be used as the primary screening method, and then use S1 recombinant protein ELISA as the re-screening method to improve the port's ability of risk perception, control and identification capabilities to mitigate the entry risk of coronavirus into the country. Bayesian statistical inference and surveillance technology based on prior distribution and optimization of port laboratory testing results can provide scientific evidence-based risk decision technology for blocking the invasion of foreign pathogenic microorganisms to the greatest extent on the basis of existing knowledge and cognition, which can be of great significance to port inspection and quarantine, sampling monitoring and risk decision-making.