R packages

('Rglasso.R')

Robust Gaussian Graphical Lasso: Estimate a precision matrix of Gaussian graphical model using lasso penalization given a robust parameter and a sparsity parameter. The computational algorithm is based on block coordinate descent algorithm.

Reference:

Sun, H. and Li, H. (2012) Robust Gaussian Graphical Modeling via l1 Penalization, Biometrics 68(4), p.1197-1206.

('Adnet.zip')

Adaptive Network-Regularized Cox Regression: Fit a adaptive network-based regularization of Cox regression with survival outcomes at a grid values of sparsity parameter lambda. A Laplacian matrix representing a graph structure among genes is imposed into a penalty function so that genetic network information can be utilized in genetic association studies with high-dimensional genomic data.

Reference:

Sun, H., Lin, W., Feng R. and Li, H. (2014) Network-Regularized High-Dimensional Cox Regression for Analysis of Genomic Data, Statistica Sinica 24(3), p.1433-1459.

(pETM)

Penalized Exponential Tilt Model: Fit a penalized exponential tilt model (ETM) to identify differentially methylated loci between cases and controls. ETM is able to detect the differences in means only, in variances only or in both means and variances. A penalized exponential tilt model using combined lasso and Laplacian penalties is applied to high-dimensional DNA methylation data from case-control association studies. When CpG sites are correlated with each others within the same gene or the same genetic region, Laplacian matrix can be imposed into the penalty function to encourage grouping effects among linked CpG sites. The selection probability of an individual CpG site is computed based on the finite number of resamplings.

Reference:

Sun, H., Wang, Y., Chen, Y., Li, Y. and Wang, S. (2017) pETM: a penalized Exponential Tilt Model for analysis of correlated high-dimensional DNA methylation data, Bioinformatics 33(12), p.1765-1772.

(pclogit)

Penalized Conditional (Unconditional) Logistic Regression: Fit a regularization path of conditional (unconditional) logistic regression model for a matched (unmatched) case-control response at a grid of values for regularization parameter lambda. A network-based penalty function was implemented to induce both sparse and smoothing variable selection. It is useful for analysis of high-dimensional data that has a specified correlation structure. Selection probability of each variable is provided.

References:

Sun, H. and Wang, S. (2012) Penalized Logistic Regression for High-dimensional DNA Methylation Data with Case-Control Studies, Bioinformatics 28(10), p.1368-1375.

Sun, H. and Wang, S. (2013) Network-based Regularization for Matched Case-Control Analysis of High-dimensional DNA Methylation Data, Statistics in Medicine 32(12), p.2127-2139.

Choi, J., Kim, K. and Sun, H.* (2018) New variable selection strategy for analysis of high-dimensional DNA methylation data, Journal of Bioinformatics and Computational Biology 16(4), 1850010.

Kim, K. and Sun, H.* (2019) Incorporating genetic networks into case-control association studies with high-dimensional DNA methylation data, BMC Bioinformatics 20:510.

Kim, K., Koo, J. and Sun, H.* (2020) An empirical threshold of selection probability for analysis of high-dimensional correlated data, Journal of Statistical Computation and Simulation 90(9) p.1606-1617

(rvsel)

Rare Variant Selection Procedure: When a gene or a genetic region is significantly associated with a disease or a trait, the rare variant selection procedure is able to distinguish causal (risk or protective) rare variants from noncausal rare variants located within the same gene or the same genetic region.

References:

Sun, H. and Wang, S. (2014) A Power Set Based Statistical Selection Procedure to Locate Susceptible Rare Variants Associated with Complex Diseases with Sequencing Data, Bioinformatics 30(16), p.2317-2323.

Kim, S., Lee, K. and Sun, H.* (2015) Statistical Selection Strategy for Risk and Protective Rare Variants Associated with Complex Traits, Journal of Computational Biology 22(11), p.1034-1043.

Lee, G. and Sun, H.* (2017) Selection Probability for Rare Variant Association Studies, Journal of Computational Biology 24(5), p.400-411.