Interactive analysis of multi-omics regulatory networks • bieulergy

Gitter GitHub issues GitHub last commit

The goal of bieulergy is to facilitate interactive network biology in R, specifically the analysis, visualization, and comparison of biological regulatory networks. This includes building out R command-line functions geared towards biological networks as well as interactive analysis through R Shiny.

Documentation

Please visit the documentation for a comprehensive overview.

Gitter

Feel free to bring your questions, comments, or feedback to our gitter channel.

Installation

We recommend the latest version of R (>= 4.0.0) and installation directly from Github.

devtools::install_github("montilab/bieulergy")

Usage

Networks are represented as interactive.omics.network objects with node/edge-level properties for mapping gene/protein symbols, centrality measures, and integrating multi-omics data layers into your analysis. These are R6 objects that extend omics.network objects to facilitate interactive analyses. You can find more information about omics.network objects here.

library(bieulergy)

Interactive Omics Network Object

library(ndexr)

ndex <- function(uuid) {
    ndexcon <- ndexr::ndex_connect()
    data <- ndexr::ndex_get_network(ndexcon, uuid)
    mat <- as.matrix(data$edges)
    storage.mode(mat) <- "character"
    ig <- igraph::graph_from_edgelist(mat[,c("s", "t")], directed=FALSE)
    ids <- data$nodes[match(as.numeric(igraph::as_ids(V(ig))), data$nodes[,"@id"]), "n"]
    V(ig)$label <- ids
    return(ig)
}

# Global landscape of HIV–human protein complexes.
# Jaeger et al. Nature. 2011 Dec 21; 481(7381): 365–370
# @UUID: 1cbe89ab-fb5d-11e9-bb65-0ac135e8bacf
# https://www.ndexbio.org/viewer/networks/1cbe89ab-fb5d-11e9-bb65-0ac135e8bacf

ig <- ndex("1cbe89ab-fb5d-11e9-bb65-0ac135e8bacf")
ionet <- bieulergy::create.ionet(ig, type="ig")

str(ionet$properties)

List of 4
 $ nodes     : int 453
 $ edges     : num 499
 $ density   : num 0.00487
 $ clustering: num 0

head(ionet$nodes)

       id label symbol degree      eigen betweenness stress
942   942   942    942     50 0.55159754  6204.11111  42589
1022 1022  1022   1022      1 0.05997156     0.00000      0
1020 1020  1020   1020      1 0.05997156     0.00000      0
 792  792   792    792      2 0.16869497    85.44444   1602
1017 1017  1017   1017      1 0.05997156     0.00000      0
1015 1015  1015   1015      1 0.05997156     0.00000      0

Graph and node properties are pre-computed for fast interactive rendering and analysis.

head(ionet$pca$var$contrib)

               Dim.1     Dim.2        Dim.3        Dim.4
degree      26.07208  4.763222 6.774052e+01  1.424178812
eigen       15.05581 84.934785 5.262416e-04  0.008874053
betweenness 28.87808  5.366217 2.387746e+01 41.878247044
stress      29.99403  4.935776 8.381493e+00 56.688700091

Multiple Network Objects

Many of the downstream functions expect one or more interactive.omics.network objects for comparative analyses.

# Simulated networks from yeast data by Kristina Hanspers
# @UUID: 7831a991-5767-11ea-bfdc-0ac135e8bacf
# https://www.ndexbio.org/viewer/networks/7831a991-5767-11ea-bfdc-0ac135e8bacf

yeast.networks <- readRDS(file.path(system.file("extdata", package="bieulergy"), "yeast-networks.rds"))

is(yeast.networks, "list")

[1] TRUE

sapply(yeast.networks, is)

                    Yeast_1                     Yeast_2                     Yeast_3 
"interactive.omics.network" "interactive.omics.network" "interactive.omics.network"

yeast.1 <- yeast.networks$Yeast_1
head(yeast.1$nodes)

     id symbol is_tf   lfc_mrna     snp_frq label degree        eigen betweenness stress
749 749   MTH1 FALSE -0.4385177 0.035319994   749      2 2.661820e-07     247.000   1361
751 751   SNF3 FALSE  0.1285503 0.003035238   751      1 5.257070e-08       0.000      0
109 109   LSM8 FALSE -0.5849400 0.062844766   109      7 1.295192e-06    3680.751  23693
743 743   ASN1 FALSE  1.1166966 0.229042875   743      1 2.299056e-04       0.000      0
692 692  SPC24 FALSE  0.2306554 0.009771783   692      3 1.164084e-03     493.000    617
740 740   GIP2 FALSE -0.5743279 0.060585153   740      2 2.958689e-03    1356.000   5772

Here is an example of plotting the top nodes by degree centrality across networks.

bieulergy::networks.tnodes(networks=yeast.networks, 
                           metric="degree",
                           symbols=TRUE,
                           size=5,
                           top=10)

Web Interface

Bieulergy is an R package and a Shiny application. Some functionality is exclusive to one or the other but there is a lot of overlap. The Shiny application is ideal for rapid analysis and comparison while the command line is more suitable for advanced custom analyses.

bieulergy::run.shiny()

Or run with Docker…

git clone https://github.com/montilab/bieulergy
cd bieulergy
docker build --tag montilab/bieulergy:latest .

Launch the package within a Docker environment..

docker run -d -p 8787:8787 -e PASSWORD=bieulergy montilab/bieulergy:latest

Run the Shiny application. Go to http://localhost:8787 in the browser and you will be promted:

username: rstudio
password: bieulergy

Then enter the following:

library(bieulergy)
bieulergy::run.shiny()

Please refer to the documentation for more examples.

bieulergy