Search

Differential network analysis investigates how the network of connected genes changes from one condition to another and has become a prevalent tool to provide a deeper and more comprehensive understanding of the molecular etiology of complex diseases. Based on the asymptotically normal estimation of large Gaussian graphical model (GGM) in the high-dimensional setting, we developed a computationally efficient test for differential network analysis through testing the equality of two precision matrices, which summarize the conditional dependence network structures of the genes. Additionally, we applied a multiple testing procedure to infer the differential network structure with false discovery rate (FDR) control. Through extensive simulation studies with different combinations of parameters including sample size, number of vertices, level of heterogeneity and graph structure, we demonstrated that our method performed much better than the current available methods in terms of accuracy and computational time. In real data analysis on lung adenocarcinoma, we revealed a differential network with 3503 nodes and 2550 edges, which consisted of 50 clusters with an FDR threshold at 0.05. Many of the top gene pairs in the differential network have been reported relevant to human cancers. Our method represents a powerful tool of network analysis for high-dimensional biological data. [ABSTRACT FROM AUTHOR]

Aiming to investigate whether genetic risk factors (GRFs) for fracture and bone mineral density (BMD) identified from people of European descent can help improve the prediction of osteoporotic fracture (OF) risk and BMD in Chinese populations, we built assessment models for femoral neck (FN)-fracture prediction and BMD value prediction using 700 elderly Chinese Han subjects and 1,620 unrelated Chinese Han subjects, respectively. 17 fracture-associated genes and 82 FN-BMD associated genes identified in people of European descent were used to build a logistic regression model with clinical risk factors (CRFs) for FN-fracture prediction in Chinese. Meanwhile 107 BMD-associated genes from people of European descent were used to build a multiple linear regression model with CRFs for BMD prediction in Chinese. A Lasso algorithm was employed for informative SNP selection to construct the genetic risk score (GRS) with ten-fold cross-validation. The results showed that, adding fracture GRF and FN-BMD GRF to the model with CRFs, the area under the receiver operating characteristic curve (AUC) decrease from 0.653 to 0.587 and 0.588, respectively, for FN fracture prediction. 62.3% and 61.8% of the risk variation were explained by the Model with CRFs and fracture GRF and by the Model with CRFs and FN-BMD GRF, respectively, as compared to 65.5% in the Model with CRFs only. The net reclassification improvement (NRI) index in the reclassification analysis is 0.56% (P = 0.57) and 1.13% (P = 0.29), respectively. There is no significant difference either between the performance of the model with CRFs and that of the model with both CRFs and GRF for BMD prediction. We concluded that, in the current study, GRF of fracture identified in people of European descent does not contributes to improve the fracture prediction in Chinese; and GRF of BMD from people of European descent cannot help improve the accuracy of the fracture prediction in Chinese perhaps partially because GRF of BMD from people of European descent may not contribute to BMD prediction in Chinese. This study highlights the limited utility of the current genetics studies largely focused on people of European descent for disease or risk factor prediction in other ethnic groups, and calls for more and larger scale studies focused on other ethnic groups. [ABSTRACT FROM AUTHOR]