Network Seriation And Modularization.
NetSAM identifies the hierarchical modules of a network (network modularization) and finds a suitable linear ordering for all leaves of the identified hierarchy (network seriation). It takes an edge-list representation of a weighted or unweighted network as input and writes files that can be used directly by the one-dimensional network visualization tool NetGestalt, or fed into other network analyses.
NetSAM can also build a correlation network (e.g. a co-expression network) from a data matrix, seriate and modularize that network, and then relate the resulting modules back to sample features or to Gene Ontology terms.
Unlike plain hierarchical clustering, NetSAM optimizes the leaf ordering, assesses the statistical significance of a network's modular organization, and identifies relevant hierarchical levels and modules at different scales.
NetSAM is distributed through Bioconductor:
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install("NetSAM")To install the development version from this repository:
BiocManager::install("bzhanglab/NetSAM")library("NetSAM")
inputNetworkDir <- system.file("extdata", "exampleNetwork.net", package = "NetSAM")
outputFileName <- file.path(getwd(), "NetSAM")
result <- NetSAM(inputNetwork = inputNetworkDir, outputFileName = outputFileName,
outputFormat = "nsm", edgeType = "unweighted",
map_to_genesymbol = FALSE, organism = "hsapiens",
minModule = 0.003, modularityThr = 0.2, nThreads = 3)| Function | Purpose |
|---|---|
NetSAM() |
Network seriation and modularization |
MatSAM() |
Correlation network construction, seriation and modularization from a matrix |
MatNet() |
Construction of a correlation network from a matrix |
consensusNet() |
Construction of a consensus co-expression network |
NetAnalyzer() |
Network analyzer |
featureAssociation() |
Calculate the associations between modules and sample features |
GOAssociation() |
Identify the associated GO terms for each module |
mapToSymbol() |
Map other ids to gene symbols |
mergeDuplicate() |
Merge duplicate ids in matrix data |
testFileFormat() |
Test whether the data matrix and annotation have a correct format |
Note that mapToSymbol() — and any function called with
map_to_genesymbol = TRUE — queries Ensembl BioMart and therefore needs a
working internet connection.
- Package vignette
- Bioconductor landing page
- From R:
browseVignettes("NetSAM")
If you use NetSAM, please cite:
Shi Z, Wang J, Zhang B. NetGestalt: integrating multidimensional omics data over biological networks. Nature Methods 10, 597–598 (2013). https://doi.org/10.1038/nmeth.2517
LGPL. Maintained by Zhiao Shi ([email protected]); originally authored by Jing Wang.