Assessing the skill of yes/no predictions

Summary Should healthy, middle‐aged women receive precautionary mammograms? Should trauma surgeons use the popular TRISS score to predict the likelihood of patient survival? These are examples of questions confronting us when we decide whether to use a yes/no prediction. In order to trust a prediction we must show that it is more valuable than would be our best guess of the future in the absence of the prediction. Calculating value means identifying our loss should the prediction err and examining the past performance of the prediction with respect to that loss. A statistical test to do this is developed. Predictions that pass this test are said to have skill. Only skillful predictions should be used. Graphical and numerical methods to identify skill will be demonstrated. The usefulness of mammograms is explored.     

Smooth tests for the zero-inflated poisson distribution

In this article we construct three smooth goodness-of-fit tests for testing for the zero-inflated Poisson (ZIP) distribution against general smooth alternatives in the sense of Neyman. We apply our tests to a data set previously claimed to be ZIP distributed, and show that the ZIP is not a good model to describe the data. At rejection of the null hypothesis of ZIP, the individual components of the test statistic, which are directly related to interpretable parameters in a smooth model, may be used to gain insight into an alternative distribution.     

Estimating mean response as a function of treatment duration in an observational study, where duration may be informatively censored

After a treatment is found to be effective in a clinical study, attention often focuses on the effect of treatment duration on outcome. Such an analysis facilitates recommendations on the most beneficial treatment duration. In many studies, the treatment duration, within certain limits, is left to the discretion of the investigators. It is often the case that treatment must be terminated prematurely due to an adverse event, in which case a recommended treatment duration is part of a policy that treats patients for a specified length of time or until a treatment-censoring event occurs, whichever comes first. Evaluating mean response for a particular treatment-duration policy from observational data is difficult due to censoring and the fact that it may not be reasonable to assume patients are prognostically similar across all treatment strategies. We propose an estimator for mean response as a function of treatment-duration policy under these conditions. The method uses potential outcomes and embodies assumptions that allow consistent estimation of the mean response. The estimator is evaluated through simulation studies and demonstrated by application to the ESPRIT infusion trial coordinated at Duke University Medical Center.     

DNA marker mining of ILSTS035 microsatellite locus on chromosome 6 of Hanwoo cattle

We describe tests for detecting and locating quantitative trait loci (QTL) for traits in Hanwoo cattle. From results of a permutation test to detect QTL for marbling, we selected the microsatellite locus ILSTS035 on chromosome 6 for further analysis. K-means clustering analysis applied to five traits and nine DNA markers in ILSTS035 resulted in three cluster groups. Finally we employed the bootstrap test method to calculate confidence intervals using the resampling method to find major DNA markers. We conclude that the major markers of ILSTS035 locus on chromosome 6 of Hanwoo cattle are markers 235 bp and 266 bp.     

重症肌无力小鼠模型的建立

目的建立重症肌无力小鼠模型。方法电鳗乙酰胆碱受体(TnAChR)和完全弗氏佐剂(CFA)混合物免疫C57BL/6J小鼠,经二次加强免疫后,检测肌力、腓肠肌肌电图、膈肌终板电镜、血清抗体水平等指标。结果与对照小鼠相比,发病小鼠表现出肌力减弱的症状,肌电图显示小鼠腓肠肌复合动作电位振幅幅度显著下降,电镜证实神经肌接头处突触后膜变平、皱褶减少、空泡样变性,发病小鼠血清抗体水平明显升高。结论成功建立了重症肌无力小鼠模型,有助于探讨其发病机理及探索治疗自身免疫病新的途径和方法。     

利用序列保守模体和局部构象信息预测转录因子结合位点

转录因子结合位点的计算预测是研究基因转录调控的重要环节,但常用的位置特异得分矩阵方法预测特异性偏低.通过深入分析结合位点的生物特征,提出了一种综合利用序列保守模体和局部构象信息的结合位点预测方法,以极大相关得分矩阵作为保守模体的描述模型,并根据二苷参数模型计算位点序列的局部构象,将两类信息得分组合为多维特征向量,在二次判别分析的框架下进行训练和滑动预测.预测过程中还引入了位置信息量以优化似然得分和过滤备选结果.针对大肠杆菌CRP和Fis结合位点数据的留一法测试结果表明,描述模型的改进和多种信息的融合能有效地改善预测方法的性能,大幅度提高特异性.     

Computation of non-covalent interactions in TNF proteins and interleukins

The roles played by the non-covalent interactions have been investigated for a set of six TNF proteins and nine Interleukins. The stabilizing residues have been identified by a consensus approach using the concepts of available surface area, medium and long-range interactions and conservation of amino acid residues. The cation-pi interactions have been computed based on a geometric approach such as distance and energy criteria. We identified an average of 1 energetically significant cation-pi interactions in every 94 residues in TNF proteins and 1 in every 62 residues in Interleukins. In TNF proteins, the cationic groups Lys preferred to be in helix while Arg preferred to be in strand regions while in Interleukins the Arg residues preferred to be in helix and Lys preferred to be in strand regions. From the available surface area calculations, we found that, almost all the cation and pi residues in TNF proteins and Interleukins were either in buried or partially buried regions and none of them in the exposed regions. Medium and long-range interactions were predominant in both TNF proteins and Interleukins. It was observed that the percentage of stabilizing centers were more in TNF proteins as compared to the Interleukins, while the percentage of conserved residues were more in Interleukins than in TNF proteins. In the stabilizing residues Lys was observed to be a stabilizing residue in both TNF proteins and Interleukins. Among the aromatic group, Phe was seen to be a stabilizing residue in both TNF and Interleukins. We suggest that this study on the computation of cation-pi interactions in TNF proteins and Interleukins would be very helpful in further understanding the structure, stability and functional similarity of these proteins.     

Correlation of protein expression, Gleason score and DNA ploidy in prostate cancer

The prognosis of prostate cancer correlates with tumor differentiation. Gleason score and DNA ploidy are two prognostic factors that correlate with prognosis. We analyzed differences in protein expression in prostate cancer of high and low aggressiveness according to these measures. From 35 prostatectomy specimens, 29 cancer samples and 10 benign samples were harvested by scraping cells from cut surfaces. DNA ploidy was assessed by image cytometry. Protein preparations from cell suspensions were examined by 2-DE. Protein spots that differed quantitatively between sample groups were identified by MS fingerprinting of tryptic fragments and MS/MS sequence analysis. We found 39 protein spots with expression levels that were raised or lowered in correlation with Gleason score and/or DNA ploidy pattern (31 overexpressed in high-malignant cancer, 8 underexpressed). Of these, 30 were identified by MS. Among overexpressed proteins were heat-shock, structural and membrane proteins and enzymes involved in gene silencing, protein synthesis/degradation, mitochondrial protein import (metaxin 2), detoxification (GST-pi) and energy metabolism. Stroma-associated proteins were generally underexpressed. The protein expression of prostate cancer correlates with tumor differentiation. Potential prognostic markers may be found among proteins that are differentially expressed and the clinical value of these should be validated.     

Ozone exposure during growth affects the feeding value of rice shoots

Rising tropospheric ozone concentrations have been observed in many Asian countries in recent years. Ozone pollution reduces the yield of agricultural crops but may also affect crop quality. This study aimed at estimating the effect of ozone exposure on feeding quality of rice shoots for ruminant herbivores. Rice plants from two genotypes differing in ozone tolerance were exposed to ozone at a concentration of 120 nl/l for 18 days, and feeding value was determined by chemical analyses and in vitro incubation in rumen fluid. Rice biomass was reduced by an average of 24% in the ozone treatment as compared to the control. Moreover, ozone exposure affected various feed quality parameters. Crude protein content was lower in ozone treated plants (P<0.05). Potential gas production during the in vitro incubation for 96 h also dropped (P<0.01) due to ozone treatment, indicating reduced digestibility of the plant materials. This was explained with an increase in the antinutritive components lignin (P<0.05) and phenolics (P<0.001) due to ozone exposure. An ozone tolerant genotype exhibited a more pronounced increase in phenolics, suggesting that this may constitute a stress defense mechanism. Our results suggest that ozone may affect the feeding value of cereal straws and calls for further research in this direction.     

Comments on the rank product method for analyzing replicated experiments

Breitling et al. [1] introduced a statistical technique, the rank product method, for detecting differentially regulated genes in replicated microarray experiments. The technique has achieved widespread acceptance and is now used more broadly, in such diverse fields as RNAi analysis, proteomics, and machine learning. In this note, we relate the rank product method to linear rank statistics and provide an alternative derivation of distribution theory attending the rank product method.