LUCID with Multiple Omics Data
Inference of LUCID model based on bootstrap resampling
Check missing patterns in one layer of omics data Z
Fit LUCID models with one or multiple omics layers
Impute missing data by optimizing the likelihood function
generate bootstrp ci (normal, basic and percentile)
I-step of LUCID
Fit a lucid model for integrated analysis on exposure, outcome and mul...
Visualize LUCID model through a Sankey diagram
Predict cluster assignment and outcome based on LUCID model using new ...
Print the output of LUCID in a nicer table
Print the output of LUCID in a nicer table
Print the output of LUCID in a nicer table
Summarize results of the early LUCID model
Summarize results of the parallel LUCID model
Summarize results of the serial LUCID model
A wrapper function to perform model selection for LUCID
An implementation of estimating the Latent Unknown Clusters By Integrating Multi-omics Data (LUCID) model (Peng (2019) <doi:10.1093/bioinformatics/btz667>). LUCID conducts integrated clustering using exposures, omics information (and outcome information as an option). This package implements three different integration strategies for multi-omics data analysis within the LUCID framework: LUCID early integration (the original LUCID model), LUCID in parallel (intermediate integration), and LUCID in serial (late integration). Automated model selection for each LUCID model is available to obtain the optimal number of latent clusters, and an integrated imputation approach is implemented to handle sporadic and list-wise missingness in multi-omics data. Lasso-type regularity for exposure and omics features were added. S3 methods for summary and plotting functions were fixed. Fixed minor bugs.