A Monte Carlo method for Bayesian analysis of linkage between single markers and quantitative trait loci. I. Methodology

A Bayesian approach to the statistical mapping of Quantitative Trait Loci (QTLs) using single markers was implemented via Markov Chain Monte Carlo (MCMC) algorithms for parameter estimation and hypothesis testing. Parameter estimators were marginal posterior means computed using a Gibbs sampler with data augmentation. Variables sampled included the augmented data (marker-QTL genotypes, polygenic effects), an indicator variable for linkage, and the parameters (allele frequency, QTL substitution effect, recombination rate, polygenic and residual variances). Several MCMC algorithms were derived for computing Bayesian tests of linkage, which consisted of the marginal posterior probability of linkage and the marginal likelihood of the QTL variance associated with the marker.     

Estimation of the position and effect of a lethal factor locus on a molecular marker linkage map

In the mapping of DNA markers the distortion of segregation of marker genotypes is often observed, which may be caused by a lethal factor acting in filial generations derived from distant crosses. A method is presented for estimating the recombination values between a lethal factor locus and neighboring molecular markers, and the relative viability or fertilization ability of gametes or zygotes affected by the lethal factor in an F₂ population using the maximum likelihood method and the expectation conditional maximization (ECM) algorithm. Three selection models of gamete or zygote were considered, and the most likely one was determined by goodness of fit of the observed frequency of the phenotypes to the expected ones under the models. The method was applied to segregation data of molecular markers of an F₂ population consisting of 144 individuals derived from a cross between an Indica and a Japonica rice variety. The presence of a lethal factor locus (L) located on chromosome III that caused partial gametic selection in both the male and female sides was suggested. The locus L was tightly linked to RFLP marker number 23 of the RFLP linkage map of Saito et al. (1991a), and the fertilization chance of a male or female gamete possessing the lethal factor was, on average, 41.5% of that of the normal gamete.     

Differences between observed and predicted energy costs at rest and during exercise in three subsistence-level populations

Estimates of daily energy expenditure are important for many areas of research in human ecology and adaptability. The most common technique for estimating human energy expenditure under field conditions, the factorial method, generally relies on activity-specific energy costs derived from published sources, based largely on North American and European subjects. There is concern that such data may not be appropriate for non-Western populations because of differences in metabolic costs. The present study addresses this concern by comparing measured vs. predicted energy costs at rest and during sub-maximal exercise in 83 subjects (52 males, 31 females) from three subsistence-level populations (Siberian herders and high-land and coastal Ecuadorian farmers). Energy costs at rest (i.e., lying, sitting and standing) and while performing a standard stepping exercise did not significantly differ among the three groups. However, resting energy costs were significantly elevated over predicted levels (+16% in men, +11% in women), whereas exercising costs were comparable to predicted values (?6% in men, +3% in women). Elevations in resting energy needs appear to reflect responses to thermal stress. These results indicate that temperature adjustments of resting energy costs are critical for accurately predicting daily energy needs among traditionally living populations. o 1996 Wiley-Liss, Inc.     

Application of radiotelemetry to the survey of acorn dispersal byApodemus mice

We examined the applicability of radiotelemetry to studies of acorn dispersal byApodemus mice and compared its efficiency with the of this spool-and-line method. Installation of a transmitter (2.2 g) onto acorns did not interfere with the transporting and feeding behavior of the mice. We were able to detect all transmitter-installed acorns and follow the daily changes in the sites in which they were hoarded, while we missed 59% of the spool-tied acorns due to mice breaking the threads. Mice carried transmitter-installed acorns farther than spool-tied ones. The radiotelemetry method is superior to the spool-and-line method and useful for the study of hoarding behavior in rodents.     

An impedimetric method for rapid screening of cosmetic preservatives

An efficient impedance method was developed for rapid evaluation of cosmetic preservatives. The method used decimal reduction time or D-value to assess preservative efficacies. The D-value, which was calculated from the plot of Log CFU ml^–1 versus time by linear regression analysis, could be obtained within 48 h. Thus, the time required for the challenge test was reduced from 4–8 weeks with the standard procedures (eg US Pharmacopeia), to 2 days with the current method. A calibration curve (r=-0.95) was established by plotting the Log CFU ml^–1 versus capacitance detection time (DT) of 108 samples. With the calibration, CFU can be estimated directly from the impedance test without plating. Two commercial biocides and several other chemicals were evaluated in a shampoo by the impedance procedure againstPseudomonas aeruginosa. The D-values obtained from the impedance test were not significantly different from those produced by the conventional plate count method. The technique was found to be particularly useful when screening a large number of compounds to find novel preservatives and synergistic preservative combinations.     

方差组分估计方法MIVQUE和REML的模拟比较

利用ＭｏｎｔｅＣａｒｌｏ方法,对４种数据结构进行了ＭＩＶＱＵＥ和ＲＥＭＬ两种方差组分估计方法的模拟比较。方差组分估计所用的模型为奶牛育种中常用的公畜模型,它包括场年季固定效应、公牛组固定效应和公牛随机效应。４种数据结构中最大的有１２８４７个观察值,场年季效应和公牛效应水平数分别为７７８和４７,它与北京市目前可利用的奶牛头胎产奶量记录资料相当。最小的数据结构只有２００个观察值,１４８个场年季和２０头公牛。比较指标为估计值的偏差和方差（理论的或根据１０００次重复模拟所得的经验值）。结果表明,对于较大样本的数据结构,两种方法差异很小,它们间的估计值的相关接近于１,偏差小于真值的１％,方差近似相等。对于较小样本的数据结构,ＭＩＶＱＵＥ则明显优于ＲＥＭＬ。本研究还表明,对于ＲＥＭＬ来说,类似数据结构１的样本已能满足其渐近无偏性和有效性的大样本特性。     

血凝法的生物载体选择与处理

用ＨＢｓＡｇＲＰＨＡ做为模式,选用羊、火鸡、鹅、人Ｏ型血球为载体,醛化后分成致敏和不致敏两种,与异嗜性抗体、中国生物制品药品检定所的参比品等作凝集试验。结果表明,无论特异性还是灵敏度都以人Ｏ型血球为最佳。在此基础上,设计了五种醛化方法处理人Ｏ型血球,共得出１８个醛化样品,经７３种不同的比较试验,筛选出最佳醛化方法为改良丙酮醛──甲醛法,采用该法醛化后血球为褐色,阴性血清和空白对照背景清晰,血球沉集点致密,下滑程度好。致敏抗—ＨＢｓＭｃＡｂ后,使ＨＢｓＡｇ灵敏度达到１～２ｎｇ／ｍｌ。为制备高质量的血凝试剂提供了最佳载体。     

不同生境区稻田节肢动物群落相似性分析

根据不同生境区稻田节肢动物群落系统调查结果,应用组平均聚类分析法及主成分分析法,讨论节肢动物群落的相似（相异）性问题。结果表明：生境条件及管理措施的差异对群落的结构影响较为明显,不同生境区的稻田节肢动物群落,都能在一定距离取值时聚合成一类,而水稻品种对此影响较小。主成分分析结果由二维图表达已基本满足要求。结果说明大八镇生境区稻田节肢动物群落与岗列镇生境区稻田节肢动物群落差异较大,分别列于两个极端。海陵镇生境区稻田介于两者之间,而与岗列镇生境区稻田节肢动物群落重叠。与聚类分析结果有所差异。这可能与主成分分析法损失部分信息有关。