The generalized Fisher's combination and accurate p-value calculation under dependence

Combining dependent tests of significance has broad applications but the related p-value calculation is challenging. For Fisher's combination test, current p-value calculation methods (eg, Brown's approximation) tend to inflate the type I error rate when the desired significance level is substantially less than 0.05. The problem could lead to significant false discoveries in big data analyses. This paper provides two main contributions. First, it presents a general family of Fisher type statistics, referred to as the GFisher, which covers many classic statistics, such as Fisher's combination, Good's statistic, Lancaster's statistic, weighted Z-score combination, and so forth. The GFisher allows a flexible weighting scheme, as well as an omnibus procedure that automatically adapts proper weights and the statistic-defining parameters to a given data. Second, the paper presents several new p-value calculation methods based on two novel ideas: moment-ratio matching and joint-distribution surrogating. Systematic simulations show that the new calculation methods are more accurate under multivariate Gaussian, and more robust under the generalized linear model and the multivariate t-distribution. The applications of the GFisher and the new p-value calculation methods are demonstrated by a gene-based single nucleotide polymorphism (SNP)-set association study. Relevant computation has been implemented to an R package GFisher available on the Comprehensive R Archive Network.     

The MASCA Approach Applied to Break Multicollinearities in Quantitative Structure-Activity Relationships

Since the regressors in multiple and multivariate regression analysis are intercorrelated, the coefficients of a regression of regressands on them are unreliable with respect to the predictive model power. Nonpredictive and predictive multicollinearity affect the ability to predict the biological activity (regressand) of novel drugs based on a function of physiochemical parameters (regressors), therefore. The nonpredictive multicollinearity can be broken by using the principal component regression analysis of the MASCA model. The working technique is exemplified on bispyridinium oxime antidotes, trimethoprim derivatives, and Taft's steric constant.     

The synthesis and biodistribution of [11C]metformin as a PET probe to study hepatobiliary transport mediated by the multi-drug and toxin extrusion transporter 1 (MATE1) in vivo

In order to develop a new positron emission tomography (PET) probe to study hepatobiliary transport mediated by the multi-drug and toxin extrusion transporter 1 (MATE1), ¹¹C-labelled metformin was synthesized and then evaluated as a PET probe. [¹¹C]Metformin ([¹¹C]4) was synthesized in three steps, from [¹¹C]methyl iodide. Evaluation by small animal PET of [¹¹C]4 showed that there was increased concentrations of [¹¹C]4 in the livers of mice pre-treated with pyrimethamine, a potential inhibitor of MATEs, inhibiting the hepatobiliary excretion of metformin. Radiometabolite analysis showed that [¹¹C]4 was not degraded in vivo during the PET scan. Biodistribution studies were undertaken and the organ distributions were extrapolated into a standard human model. In conclusion, [¹¹C]4 may be useful as a PET probe to non-invasively study the in vivo function of hepatobiliary transport and drug–drug interactions, mediated by MATE1 in future clinical investigations.     

基于核范数正则化的抗癌药物组合协同作用预测

目的 抗癌药物联合疗法是一种很有前途的治疗策略。针对特定癌症类型，选择高度协同的药物组合，对提高癌症疗效至关重要。然而，确定具有协同作用的药物组合是一项复杂而困难的工作。本研究旨在完全以数据驱动、计算建模的方式优化抗癌药物组合高通量虚拟筛选，为“旧药重新定位新组合”提供理论参考。方法 借鉴矩阵填充思想，构建了基于核范数正则化的计算模型NNRM，用于预测抗癌药物组合的协同得分和协同状态。针对固定细胞系构造对称的协同得分观测矩阵；采用分折技巧将观测矩阵稀疏化；借助“交替方向乘子法”和“软阈值估计”求解模型。结果 将NNRM应用于O’Neil团队发布的数据集，预测的协同得分与观测值之间的均方根误差为14.78，预测的协同状态准确率为0.94，优于随机森林（RF）和支持向量机（SVM），完全可以与深度学习模型相媲美。此外，NNRM预测的部分缺失值结果与已有研究或临床实践相吻合。结论 NNRM可实现大规模、批量预测抗癌药物组合的协同作用，极大地降低了已有模型对数据的要求和计算成本，缩短了高通量虚拟筛选的测试时间，可以作为抗癌药物组合高通量虚拟筛选的可选择工具。     

Oleanolic acid increases the anticancer potency of doxorubicin in pancreatic cancer cells

Combination therapy is a novel cancer therapy approach that combines two or more chemotherapy drugs. This treatment modality enhances the efficacy of chemotherapy by targeting key pathways in an additive or synergistic manner. Therefore, we investigated the efficacy of combination therapy by widely used chemotherapy drug doxorubicin (DOX) and oleanolic acid (OA) to induction of apoptosis for pancreatic cancer (PC) therapy. The effects of DOX, OA, and their combination (DOX-OA) were investigated on proliferation and viability of PC cell line (PANC-1) by MTT assay. Moreover, migration and invasion of the cancer cells were evaluated by trans-well migration assay and wound healing assay. Flow cytometry and DAPI (4′,6-diamidino-2-phenylindole) staining were employed to investigate apoptosis quantification and qualification of the treated cancer cells. Finally, mRNA expression of apoptosis-related genes was assessed by quantitative real-time polymerase chain reaction. Our results demonstrated that the proliferation and metastasis potential of PC cells significantly decreased after treatment by DOX, OA, and DOX-OA. Moreover, we observed an increase in apoptosis percentage in the treated cancer cells. The apoptosis-related gene expression was modified to increase the apoptosis rate in all of the treatment groups. However, the anticancer potency of DOX-OA combination was significantly more than that of DOX and OA treatments alone. Our study suggested that DOX-OA combination exerts more profound anticancer effects against PC cell lines than DOX or OA monotherapy. This approach may increase the efficiency of chemotherapy and reduce unintended side effects by lowering the prescribed dose of DOX.     

Landslide Hazard and Risk Mapping Using the Weighted Linear Combination Model Applied to the Tevankarai Stream Watershed,Kodaikkanal, India

Natural hazards like landslides, earthquakes, and floods are a major deterrent to the development of mountain regions of the world. Recently, there has been a rise in the number of landslides in Western Ghats, India. Several factors cause landslides and they depend on the local geo-environmental set-up of the region. This study attempts to map the spatial distribution of landslide hazard and analyze the risk for a typical hill town, Kodaikkanal in the Western Ghats, India, facing rapid urbanization and infra-structure growth, using a weighted linear combination model in a geographic information system platform. Landslides in the region are triggered by rainfall and it is fairly uniform throughout the area. Hence, the susceptibility map is used as the hazard map. Validation of the weighted linear combination model using landslide hazard index shows that landslide density increases with the hazard class. The risk assessment matrix (RAM) is used to evaluate risk based on the land use and landslide hazard category. The land use map is reclassified by assigning damage potential for each land use feature. The risk map is classified into low, moderate, and high risk categories. Suitable control measures are suggested for various risk categories.     

The Heliotropium marifolium complex in India and five new combinations in Euploca (Heliotropiaceae) from H. sect. Orthostachys (Boraginaceae s.l.)

This paper includes studies on Heliotropium marifolium complex from India. Five new combinations in Euploca from H. sect. Orthostachys are proposed here. H. marifolium, H. marifolium subsp. wallichii, H. rottleri, H. cornutum and H. madurense are transferred to the genus Euploca. The taxonomic status of the unresolved name H. rottleri is also discussed.