A Tool for 'Covariate'-Sensitive Longitudinal Analysis on 'omics' Data
Effect size (Cliff's delta) calculation in longdat_disc() pipeline
Covariate model test in longdat_cont() pipeline
Covariate model test in longdat_disc() pipeline
Post-hoc test based on correlation test for longdat_cont().
Create cuneiform plots of result table from longdat_disc() or longdat_...
Data preprocessing
Calculate the p values for every factor (used for selecting factors la...
Generate result table as output in longdat_cont()
Generate result table as output in longdat_disc()
Replace the symbols in variable and covariate names in raw input
Longitudinal analysis with time as continuous variable
Longitudinal analysis with time as discrete variable
Create input master table from metadata and feature tables for longdat...
Null Model Test and post-hoc Test in longdat_cont() pipeline
Null Model Test and post-hoc Test in longdat_disc() pipeline
Randomized negative control for count data in longdat_cont()
Randomized negative control for count data in longdat_disc()
Remove the dependent variables that are below the threshold of sparsit...
Remove the dependent variables that are below the threshold of sparsit...
Plot theta values of negative binomial models versus non-zero count fo...
Unlist confound (covariate) and inverse confound (covariate) tables, t...
Wilcoxon post-hoc test
This tool takes longitudinal dataset as input and analyzes if there is significant change of the features over time (a proxy for treatments), while detects and controls for 'covariates' simultaneously. 'LongDat' is able to take in several data types as input, including count, proportion, binary, ordinal and continuous data. The output table contains p values, effect sizes and 'covariates' of each feature, making the downstream analysis easy.