Bayesian Networks PCfdr Algorithm [For download, please send request email to junningl@gmail.com] The PCfdr algorithm is a structure-learning algorithm that controls the false discovery rate (FDR) of network connections inferred from data. It allows users to specify the upper bound of the desired FDR in advance and then it learns the network structure with the FDR controlled accordingly.
(1) Controlling the False Discovery Rate of the Association/Causality Structure Learned with the PC Algorithm (2) Learning brain connectivity with the false-discovery-rate-controlled PC-algorithm Junning Li, Z. Jane Wang and Martin J. McKeown Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 4617 - 4620, 2008. [url] Multiple Testing Meff [online service] Meff is algorithms for estimating the effective number of independent tests in correlated multiple testing. It has been widely used to control the experiment-wise error rate in genetic association studies. It can provide results comparable with the permutation test, but needs much less computation. Disclaimer and Acknowledgment The idea of the Meff method was first introduced by Dr. James M. Cheverud, and later I proposed a variation which significantly improves its accuracy. I did not implement the Meff service linked here. It was by Dr. Dale R. Nyholt, to whom we owe thanks for his implementing and providing the online service. Recently, a few new improvements and variations of the Meff method have been published, but may have not been included in the online service. Tags: multiple hypothesis testing, multilocus analysis, association study, experiment-wise error rate References (in inverse chronological order): (1) Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix (2) A simple correction for multiple testing for SNPs in linkage disequilibrium with each other Dale R. Nyholt Am. J. Hum. Genet., 74: 765 - 769, 2004. |