dynUGENE is an R package for the inference, simulation, and visualization of gene regulatory network dynamics from time-series expression data.

dynUGENE build off dynGENIE3, an algorithm to infer gene network architecture and dynamics given time-series or steady-state expression data.

dynUGENE provides several additional functionalities on top of dynGENIE3.

• Visualization of the inferred network as a heatmap.
• Simulation of the learned system given any initial condition.
• Stochastic simulations by accounting for uncertainty in the random forests’ predictions
• Model selection using a Pareto front by comparing model error with model complexity.
• Additional datasets (repressilator, Hodgkin-Huxley) both deterministic and stochastic.

The package is useful for those who wish to build explainable models of gene regulatory networks with varying degrees of complexity. The package is also useful for synthetic biologists who wish to design genetic circuits that match some desired dynamical or steady-state properties.

Install the development version with:

install.packages("devtools")
devtools::install_github("tianyu-lu/dynUGENE", build_vignettes = TRUE)
library("dynUGENE")

The package is also available as a Shiny app on at tianyulu.shinyapps.io/dynUGENE/ or can be run locally with

dynUGENE::rundynUGENE()

The Shiny app includes a tutorial and interactivity with the outputs of dynUGENE. In particular, useful features of the app include the following:

• Visualize each of the stepwise-tuned networks
• Select custom masks by clicking on the heatmap cells to be masked
• Interactive simulation plot allows zooming

library("dynUGENE")
ls("package:dynUGENE")
#>  [1] "estimateDecayRates"      "HodgkinHuxley"          
#>  [3] "inferNetwork"            "inferSSNetwork"         
#>  [5] "plotTrajectory"          "Repressilator"          
#>  [7] "rundynUGENE"             "simulateUGENE"          
#>  [9] "StochasticHodgkinHuxley" "StochasticRepressilator"
#> [11] "tuneThreshold"
data(package="dynUGENE")

The package file structure is illustrated below.

For detailed tutorials and descriptions of the provided datasets, see the vignette here:

browseVignettes("dynUGENE")
#> starting httpd help server ... done

Package author: Tianyu Lu.

With the exception of estimateDecayRates.R, the remaining files are original code. inferNetwork.R is a re-implementation of the dynGENIE3 algorithm in R. The original implementation wraps around a random forest implementation written in C. dynuGENE implements everything in R. The randomForest package is used to train random forests. ramify is used to obtain column-wise argmax of matrices. Plots for inferNetwork() are made with ggplot2 and preprocessing done by reshape2. Plots for inferSSNetwork() are made with gplots and colours specified by RColorBrewer. The stats package is used to estimate mean and variance from the random forests predictions and to sample from a Gaussian.

Sources for code adapted from examples are provided near the code in question.

We can learn the architecture and simulate the dynamics of a repressilator with inferNetwork():

We can perform a search over possible sparse network architectures with tuneThreshold().

We can also learn the dynamics of a stochastic repressilator:

We can also visualize the learned networks for large, steady-state datasets:

Load and do inference with repressilator data:

library(dynUGENE)

## Infer network
ugene <- inferNetwork(Repressilator, mtry = 3L)

## Deterministic simulation of the inferred network dynamics
x0 <- Repressilator[1, 2:7]
trajectory <- simulateUGENE(ugene, x0)
plotTrajectory(trajectory, c("p3", "p2", "p1"))

## Stochastic simulation
trajectory <- simulateUGENE(ugene, x0, stochastic = TRUE)
plotTrajectory(trajectory, c("p3", "p2", "p1"))

inferNetwork() take about 30 seconds for automatic threshold tuning for the Repressilator dataset.

simulateUGENE() takes about 3 minutes for simulating 5000 timesteps of a 6 component system.

Tests: running the unit and integration tests takes about three minutes on a typical laptop.

dynUGENE welcomes issues and other contributions.

Logo made with Wix.

Elowitz, M. B., & Leibler, S. (2000). A synthetic oscillatory network of transcriptional regulators. Nature, 403(6767), 335-338.

Geurts, P. (2018). dynGENIE3: dynamical GENIE3 for the inference of gene networks from time series expression data. Scientific reports, 8(1), 1-12.

Pau Bellot, Catharina Olsen and Patrick E Meyer (2020). grndata: Synthetic Expression Data for Gene Regulatory Network Inference. R package version 1.20.0.

Hodgkin, A. L., & Huxley, A. F. (1952). A quantitative description of membrane current and its application to conduction and excitation in nerve. The Journal of physiology, 117(4), 500.

Mangan, N. M., Brunton, S. L., Proctor, J. L., & Kutz, J. N. (2016). Inferring biological networks by sparse identification of nonlinear dynamics. IEEE Transactions on Molecular, Biological and Multi-Scale Communications, 2(1), 52-63.

A. Liaw and M. Wiener (2002). Classification and Regression by randomForest. R News 2(3), 18–22.

BioRender. (2020). Image created by Lu, T. Retrieved December 2, 2020, from https://app.biorender.com/

Brandon Greenwell (2016). ramify: Additional Matrix Functionality. R package version 0.3.3. https://CRAN.R-project.org/package=ramify

Brewer, C., Harrower, M., Sheesley, B., Woodruff, A., & Heyman, D. (2020). Colorbrewer 2.0. Retrieved December 02, 2020, from https://colorbrewer2.org/

Carl Ganz (2016). rintrojs: A Wrapper for the Intro.js Library. Journal of Open Source Software, 1(6), October 2016. URL http://dx.doi.org/10.21105/joss.00063

Erich Neuwirth (2014). RColorBrewer: ColorBrewer Palettes. R package version 1.1-2. https://CRAN.R-project.org/package=RColorBrewer

Ggplot2 : Quick correlation matrix heatmap - R software and data visualization. (2020). Retrieved December 02, 2020, from http://www.sthda.com/english/wiki/ggplot2-quick-correlation-matrix-heatmap-r-software-and-data-visualization

Gregory R. Warnes, Ben Bolker, Lodewijk Bonebakker, Robert Gentleman, Wolfgang Huber, Andy Liaw, Thomas Lumley, Martin Maechler, Arni Magnusson, Steffen Moeller, Marc Schwartz and Bill Venables (2020). gplots: Various R Programming Tools for Plotting Data. R package version 3.1.0. https://CRAN.R-project.org/package=gplots

H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.

Hadley Wickham and Jennifer Bryan (2020). usethis: Automate Package and Project Setup. R package version 1.6.3. https://CRAN.R-project.org/package=usethis

Hadley Wickham, Jim Hester and Winston Chang (2020). devtools: Tools to Make Developing R Packages Easier. R package version 2.3.2. https://CRAN.R-project.org/package=devtools

Hadley Wickham, Peter Danenberg, Gábor Csárdi and Manuel Eugster (2020). roxygen2: In-Line Documentation for R. R package version 7.1.1. https://CRAN.R-project.org/package=roxygen2

Hadley Wickham (2007). Reshaping Data with the reshape Package. Journal of Statistical Software, 21(12), 1-20. URL http://www.jstatsoft.org/v21/i12/.

Hadley Wickham. testthat: Get Started with Testing. The R Journal, vol. 3, no. 1, pp. 5–10, 2011

R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.

Rackauckas, C., & Nie, Q. (2017). Adaptive methods for stochastic differential equations via natural embeddings and rejection sampling with memory. Discrete and continuous dynamical systems. Series B, 22(7), 2731.

Rackauckas, C., & Nie, Q. (2017). Differentialequations. jl–a performant and feature-rich ecosystem for solving differential equations in julia. Journal of Open Research Software, 5(1).

Silva, A. (2020) TestingPackage: An Example R Package For BCB410H. Unpublished. URL https://github.com/anjalisilva/TestingPackage

Wickham, H. and Bryan, J. (2019). R Packages (2nd edition). Newton, Massachusetts: O’Reilly Media.

Winston Chang, Joe Cheng, JJ Allaire, Yihui Xie and Jonathan McPherson (2020). shiny: Web Application Framework for R. R package version 1.5.0. https://CRAN.R-project.org/package=shiny

Yihui Xie (2020). knitr: A General-Purpose Package for Dynamic Report Generation in R. R package version 1.30.