Most of my open-source software is available for download at my GitHub site: here


blockRAR is a tool for the simulation of response-adaptive randomization trials. The package supports binomial outcomes. To adjust for time trends in clinical trials, patients are assigned in block and randomization ratio is altered by block. Unlike traditional RAR where the randomization ratio is altered patient by patient basis. The package allow simulations using both frequentist and Bayesian approaches. [Vignette] [CRAN] [Source] [Webpage]


bayesCT is a tool for simulation for adaptive Bayesian clinical trials. It is designed to simulate and analyze adaptive Bayesian trials for different parameters and distributions. The bayesian adaptive trial accounts for interim looks and computes power and type 1 error. It also checks for futility and early stopping for success. Monte Carlo simulation is used for imputation if there is missing data in the interim looks. bayesCT allows incorporation of historical data in the simulation if provided. bayesCT also supports imbalanced randomization. It is still under development. [Vignette] [CRAN] [Source] [Webpage] [Shiny App]


Lag Penalized Weighted Correlation (LPWC) is a method for clustering short time series data. It is designed to identify groups of biological entities (for example, genes or phosphosites) that exhibit the same pattern of activity changes over time. LPWC allows lags to incorporate delayed responses in the biological data. For example, two genes may have similar expression changes over time, but one initiates those changes 5 minutes after the other. LPWC also supports irregular time intervals between time points collected in biological data. [Vignette] [CRAN] [Source] [Webpage]


Sample size for non-inferiority trial(SSNI) computes sample size for non-inferiority trial with additive and multiplicative margin. It is designed to compute the sample size allocated for both treatment and control group. The efficiency of optimal allocation compared to (1:1) equal allocation is reported. [App] [Code]