Package: dicepro 1.0.2

Kalidou BA

dicepro: Semi-Supervised Deconvolution of Bulk RNA-Seq Data with Hyperparameter Optimization

Performs semi-supervised deconvolution of bulk RNA sequencing (RNA-seq) data. Known cell-type proportions are estimated using supervised methods -- 'CIBERSORTx' (CSx), 'CIBERSORT' (CS), 'FARDEEP' (Fast And Robust DEconvolution of Expression Profiles), and 'DCQ' (Digital Cell Quantifier) -- while unknown components are inferred using non-negative matrix factorization ('NMF') with limited-memory Broyden-Fletcher-Goldfarb-Shanno with bounds ('L-BFGS-B') optimization. Hyperparameters are selected automatically using a Pareto-frontier-based approach with knee-point detection, allowing application when the reference signature matrix is incomplete. More details about 'DICEpro' can be found in Ba et al. (2026) <doi:10.64898/2026.06.17.732876>.

Authors:Kalidou BA [aut, cre], Boris P. Hejblum [aut]

dicepro_1.0.2.tar.gz
dicepro_1.0.2.tar.gz(r-4.7-arm64)dicepro_1.0.2.tar.gz(r-4.7-x86_64)dicepro_1.0.2.tar.gz(r-4.6-arm64)dicepro_1.0.2.tar.gz(r-4.6-x86_64)
dicepro_1.0.2.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
dicepro/json (API)

# Install 'dicepro' in R:
install.packages('dicepro', repos = c('https://cran.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/kalidouba/dicepro/issues

Uses libs:
  • c++– GNU Standard C++ Library v3
  • openmp– GCC OpenMP (GOMP) support library
Datasets:

On CRAN:

Conda:

cppopenmp

3.18 score 7 scripts 23 exports 56 dependencies

Last updated from:98e5ba8f5f. Checks:6 OK. Indexed: yes.

TargetResultTimeFilesSyslog
linux-devel-arm64OK210
linux-devel-x86_64OK226
source / vignettesOK221
linux-release-arm64OK212
linux-release-x86_64OK180
wasm-releaseOK180

Exports:best_hyperParamscontains_nan_or_infcreate_gamma_lambda_plotdiceprofull_metricsgenerate_ref_matrixgeneratePropheatmap_abundancesmakeTable1Toolmetric_plotnmf_lbfgsbnmf_lbfgsb_hyperOptnrmseobjective_optplot_hyperoptresearch_hyperOptrow_norm_posrun_CSxrun_experimentrunning_methodsamplewise_metricssimulationsimulation_bluecode

Dependencies:bootclassclicodetoolsComICScpp11dplyre1071FARDEEPfarverforeachgenericsggplot2glmnetgluegtableisobanditeratorsKraljicMatrixlabelinglatticelbfgsb3clifecyclelme4magrittrMASSMatrixminqanlmenloptrnnlsnumDerivpatchworkpillarpkgconfigpreprocessCoreproxyR6rbibutilsRColorBrewerRcppRcppArmadilloRcppEigenRdpackreformulasrlangS7scalesshapesurvivaltibbletidyselectutf8vctrsviridisLitewithr

dicepro - Hyperparameter Search Space Visualization
Overview | The Two Strategies | "all" - Independent sampling | "restrictionEspace" - Linked sampling | Visualizing the Search Space | "all" - Independent space | "restrictionEspace" - Restricted space | Simulated Data | Running the optimization | Strategy "all" - Independent sampling | Strategy "restrictionEspace" - linked sampling | Comparing the Two Strategies | Session Info
Rendered fromhyperparameter-space.Rmdusingknitr::rmarkdown

Last update: 2026-06-30
Started: 2026-06-30

dicepro - Real Data Workflow (BlueCode + CellMixtures)
Overview | Data Loading | BlueCode Reference Signature Matrix | CellMixtures Bulk Dataset | Data Inspection | Gene Overlap | Expression Distribution | Deconvolution with dicepro() | Results | Optimal Hyperparameters | Hyperparameter Optimization Report | Pareto Frontier | Estimated Cell-Type Proportions | Top Contributing Cell Types | Per-Sample Composition | Compartment-Level Summary | Session Info
Rendered fromvignette-real-data.Rmdusingknitr::rmarkdown

Last update: 2026-06-30
Started: 2026-06-30

dicepro - Simulated Data Workflow
Overview | Strategy 1 -- Fully Synthetic Simulation | Data Generation | Noise Model Sanity Check | Deconvolution | Results | Optimal Hyperparameters | Hyperparameter Optimisation Report | Pareto Frontier | Recovered vs True Proportions | Per-Cell-Type Correlation | Quantitative Performance Metrics | Strategy 2 -- BlueCode-Based Simulation | Compartment Structure | Proportion Distribution by Compartment | Comparing Both Strategies | Session Info
Rendered fromvignette-simulation.Rmdusingknitr::rmarkdown

Last update: 2026-06-30
Started: 2026-06-30

Readme and manuals

Help Manual

Help pageTopics
dicepro: Semi-Supervised Deconvolution of Bulk RNA-Seq Data with Hyperparameter Optimizationdicepro-package
Select optimal hyper-parameters using a Pareto frontierbest_hyperParams
BlueCode Reference Signature MatrixBlueCode
CellMixtures Bulk RNA-seq DatasetCellMixtures
Check if a value contains NaN or Infcontains_nan_or_inf
Visualize the gamma–lambda hyper-parameter search spacecreate_gamma_lambda_plot
Semi-supervised bulk RNA-seq deconvolution with hyper-parameter optimizationdicepro
Full agreement metrics via mixed-effects modelingfull_metrics
Generate a Reference Signature Matrixgenerate_ref_matrix
Generate Cell-Type Proportion MatrixgenerateProp
Heatmap of cell-type abundancesheatmap_abundances
Build a performance metrics table for composition matricesmakeTable1Tool
Performance metric line plot across iterationsmetric_plot
NMF with L-BFGS-B optimizationnmf_lbfgsb
NMF L-BFGS-B wrapper for hyper-parameter optimizationnmf_lbfgsb_hyperOpt
Normalized Root Mean Square Error (NRMSE)nrmse
Objective function for hyper-parameter optimizationobjective_opt
Plot hyperparameter optimization reportplot_hyperopt plot_hyperopt.dicepro
Plot cell abundance heatmap and error plotplot.dicepro
Hyper-parameter optimization loop for diceproresearch_hyperOpt
Row-normalize a matrix with non-negative clampingrow_norm_pos
Run CIBERSORTx Deconvolution Methodrun_CSx
Run a dicepro hyperparameter optimization experimentrun_experiment
Run cell-type deconvolutionrunning_method
Sample-wise Pearson correlation and RMSEsamplewise_metrics
Simulate Bulk RNA-seq Data with Biological and Technical Noisesimulation
Simulate Bulk RNA-seq Data Using the BlueCode Reference Matrixsimulation_bluecode