If you want to create manually a new file to use with GRNsight, format the file in the following manner:

• GRNsight accepts Microsoft Excel Workbook (*.xlsx) files as its native format.
  • Note that Excel 97-2003 Workbook (*.xls) files are not able to be read by GRNsight.
  • If you have a SIF (*.sif) or GraphML (*.graphml) file, please see Section 1b-i.c or Section 1b-i.d, respectively, to verify that your files are formatted correctly for import into GRNsight.
• GRNsight reads gene regulatory network information that is in the form of an adjacency matrix to generate the nodes and edges of the graph.
  • There must be one worksheet in your Excel Workbook entitled “network” or “network_optimized_weights”. Generally, the “network” worksheet describes an unweighted network and the “network_optimized_weights” worksheet describes a weighted network. If the Workbook contains worksheets with both names, the display of the “network_optimized_weights” data will take precedence.
  • In the “network” and “network_optimized_weights” worksheets, the upper-left cell (A1) should contain the text “cols regulators/rows targets”. This text is there as a reminder of the direction of the regulatory relationships specified by the adjacency matrix.
  • The rest of the first row should contain the names of the transcription factors that are controlling the other genes in the network, one transcription factor name per column. Transcription factor names must be unique for each column and must be 12 characters or fewer.
  • The rest of the first column should contain the names of the target genes that are being controlled by the transcription factors heading each of the columns in the matrix, one target gene name per row. Transcription factor names must be unique for each row and must be 12 characters or fewer.
  • GRNsight is primarily designed to depict gene regulatory networks that contain just transcription factors and that can be represented by a symmetric adjacency matrix. However, starting with version 1.8, GRNsight allows the user to upload spreadsheets with asymmetric adjacency matrices. I.e., the number, identity, and order of genes across the top and side of the adjacency matrix do not have to match. A gene name at the top of the matrix will be considered the same as a gene name on the side if it contains the same text string, regardless of capitalization. For example, “ABF1”, “Abf1”, and “abf1” would all be considered the same gene, but “ABF_1” or “ABF-1” would not be considered the same as “ABF1”. Gene names may not contain any special characters other than "-" and "_". GRNsight is case-preserving in that it will keep the capitalization that the user originally entered into the spreadsheet. The gene name capitalization occurring across the top takes precedence over the gene name capitalization occurring on the side of the adjacency matrix for display of the gene names on the nodes of the graph. The nodes that display on the graph are automatically sized to accommodate the size of the label. Labels are restricted to 12 characters; use shorter gene symbols for nodes so that the node boxes do not get too large.
  • For unweighted networks in the “network” worksheet, each cell in the matrix should then contain a zero (0) if there is no regulatory relationship between those two transcription factors, or a one (1) if there is a regulatory relationship between them. In the screenshot below, the “1” in cell B7 means that ABF1 regulates FHl1. The “1” in cell D4 means that AFT1 regulates itself. When this file is loaded into GRNsight, it will generate a graph with 21 nodes, one for each transcription factor, and 31 edges, for each of the cells that contain a “1” in this adjacency matrix. The edges will be black with regular (pointy) arrowheads. The demonstration file "Demo #3: Unweighted GRN" corresponds to the adjacency matrix shown below. Follow the link to download the file.
    
    
    
    
  • When GRNsight encounters a worksheet entitled “network” with all 1’s and 0’s specifying an adjacency matrix, it will automatically color the edges black and put regular (pointy) arrowheads on the edges. However, GRNsight can also be used to display weighted edges with varying thicknesses and colors based on the magnitude and sign of the weights, respectively. To generate this type of adjacency matrix by hand, create a worksheet called “network_optimized_weights”. Instead of using 1’s to indicate connections in the network, use values that are real numbers > 0 to indicate an activation relationship or < 0 to indicate a repression relationship. The screenshot below gives an example of this type of adjacency matrix. The demonstration file Demo #4: Weighted GRN corresponds to the adjacency matrix shown below. Follow the link to download the file. For the weighted edges to display with different colors and thicknesses, make sure that you have selected the menu item Edit > Preferences > Format edges based on optimized weight parameters. For more information on how GRNsight displays weighted edges on the graph, see Section 2 below.
    
    
    
    
  • Note that GRNsight is designed to visualize “small-scale” or “medium-scale” gene regulatory networks, not the entire gene regulatory network for an organism. Currently, it is recommended that you upload networks with no more than 35 unique genes or 70 edges. If you upload a network with 50 nodes or 100 edges or more, you will receive a warning and the network graph will still display. If you attempt to upload a network of 75 or more nodes or 100 or more edges, the graph will not display.
  • Follow these instructions to download and format a regulation matrix from the YEASTRACT database for use with GRNsight.
• Starting in v3.0.0, GRNsight reads time series expression data in the form of an adjacency matrix, and displays this data as a node coloring visualization.
  • Excel Workbook with names ending with "_expression" are processed as expression time series data. The upper-left cell (A1) should contain the text “id”.
  • The rest of the first row should contain the names of the transcription factors under test, one transcription factor name per column. Transcription factor names must be unique for each column and must be 12 characters or fewer.
  • The rest of the first column should contain numerical time points in ascending order.
  • The values withing the adjancy matrix should represent the expression value of the particular transcription factor at a given time point. Blank cells are allowed, indicating the lack of a data point for a particular transcription factor at a given time.
  • Please see Section 2e for more information on how to interpret and manipulate the node coloring visualization.
• Starting in v4.0.0 GRNsight reads metadata from a sheet entitled “optimization_parameters.”
  • The header should contain “optimization” cell A1 and “value” in cell B1.
  • The following optimization parameters can be a number: “alpha”, “kk_max”, “MaxIter”, “TolFun”, “MaxFunEval”, “TolX”, “L_curve”, “estimate_params”, “make_graphs”, “fix_P”, “fix_b”, “taxon_id”, and “b_or_tau”.
  • The following optimization parameters can be a string: “production_function” and “species”.
  • The following optimization parameters can be a list of numbers: “expression_timepoints” and “simulation_timepoints”.
  • The following optimization parameters can be a list of numbers: “expression_timepoints” and “simulation_timepoints”.

This section has instructions for formatting a gene regulation matrix downloaded from YEASTRACT as a .xlsx file in the native GRNsight format.

• Navigate to Generate Regulation Matrix on the YEASTRACT site.
• Select the appropriate radio buttons/check boxes for the filters.
• Paste a list of transcription factors into the appropriate field.
• Paste a list of targets into the Target ORF/Genes field, or check the box to consider all ORF/Genes.
• Click the Generate button.
• In the results window that appears, click the link to download the Regulation matrix results file as a Semicolon Separated Values (CSV) file (not the Two-column table, Tab Separated Values (TSV) file).
• Once you have downloaded the file, launch Microsoft Excel.
• Select the menu item, File > Open and select the file that you downloaded.
• Select Column A.
• Select the menu item, Data > Text to Columns...
• In the first window of the wizard that appears, select the radio button for "Delimited" and click Next.
• In the second window of the wizard that appears, check the box for "Other" under "Delimiters" and type a semicolon in the field to the right and click Finish.
• Select the menu item, File > Save As... and save the file as an Excel Workbook (.xlsx).
• YEASTRACT organizes the adjacency matrix with the transcription factors as rows and the target genes as columns, which is opposite to the format that GRNmap and GRNsight expects. To flip this orientation, first create a new worksheet by clicking on the new worksheet icon at the bottom of the screen. Name this new worksheet "network".
• Select the adjacency matrix from the first worksheet and copy it to the clipboard. Go to the "network" worksheet and click on cell A1. Select the menu item Edit > Paste Special. In the window that appears, check the box "Transpose" and click OK.
• The labels for the genes in the columns and rows needs to match. Thus, delete the "p" from each of the gene names in the columns.
• Paste the following text into cell A1 "cols regulators/rows targets".
• Save your work. This file is now ready for loading into GRNsight. You can choose to delete the original worksheet or keep it in your workbook.

Simple Interaction Format (SIF or .sif) was originally created for use with Cytoscape, an open source bioinformatics software platform for visualizing molecular interaction networks, but is now used to exchange network data between many different software packages. Users who have data in SIF format can import these files directly into GRNsight. However, if you are creating a new input file for use with GRNsight, we recommend using the native Excel format expected by GRNsight as described in Section 1a. An extended description of the SIF format can be found in the Cytoscape v3.4.0 User Manual, from which the description below is derived.

• A SIF file is a tab-delimited text file with the file extension .sif.
• Note that the Cytoscape SIF specification also allows space-delimited files. However, due to the confusion caused when a space character is used in a node label or interaction type, GRNsight only allows tabs as delimiters for .sif files.
• Lines in the SIF file specify a source node, a relationship type (or edge type), and one or more target nodes separated by tab characters.
• For a network defined in SIF format, node names should be unique, as identically named nodes will be treated as identical nodes.
• Node names must be 12 characters or fewer.
• Starting from GRNsight v2, the relationship type for a weighted network is required to be "pd". "pd" was chosen because it commonly means protein -> DNA (e.g. a regulatory transcription factor binding upstream of a target gene) in the systems biology community. A strict requirement for the relationship type was specified to expand upon SIF syntactic error checking capabilities.
• Duplicate entries are ignored (as would occur if the same source-target pair was listed twice with two different relationship types).
• GRNsight automatically detects whether the SIF file represents an unweighted or weighted network through the examination of the relationship type. The screenshots and descriptions below are four valid ways that an unweighted network can be specified as a SIF file, using the file Demo #3: Unweighted GRN" as an example. (The screenshots display a tab-delimited text file as it would be displayed in Microsoft Excel.) Follow the link to download a sample SIF file.
• In the first example below, each source node in the network is listed in the first column, the relationship type “pd” is listed in the second column, and the target node is listed in the third column. If a node has multiple target nodes, it is listed multiple times, one relationship per line. A node may regulate itself (for example, AFT1). Note that FHL1, GTS1, MSN1, and RPH1 do not have any target nodes, but are still listed as source nodes. While they can be listed, it is not necessary to do so, as is shown in the second example.
  
  
  
  
• In the second example below, only source nodes that have target nodes are listed in the first column, the relationship type “pd” is listed in the second column, and the target node is listed in the third column. If a node has multiple target nodes, it is listed multiple times, one relationship per line. (FHL1, GTS1, MSN1, and RPH1 were removed because they did not have any target nodes.)
  
  
  
  
• In the third example below, each source node is only listed once on one line, followed by the relationship type “pd”, followed by a series of target nodes, separated by tab characters. Note that FHL1, GTS1, MSN1, and RPH1 do not have any target nodes, but are still listed as source nodes. While they can be listed, it is not necessary to do so, as is shown in the fourth example.
  
  
  
  
• In the fourth example below, each source node is only listed on one line, followed by the relationship type “pd”, followed by a series of target nodes, separated by tab characters. (FHL1, GTS1, MSN1, and RPH1 were removed because they did not have any target nodes.)
  
  
  
  
• A Cytoscape SIF file is only intended to convey the network nodes and edge connections; it does not support the storage of other node or edge attributes or display properties. However, with a modification to the relationship type, a SIF file can represent a weighted network and be displayed as such in GRNsight upon import (See Section 2a).
• To encode a weighted network in a SIF file, instead of using a text string like “pd” as the relationship type, substitute the numerical weight value. Use real numbers > 0 to indicate an activation relationship or < 0 to indicate a repression relationship. The screenshots and descriptions below are two valid ways that a weighted network can be specified as a SIF file, using the file Demo #4: Weighted GRN" as an example. (The screenshots display a tab-delimited text file as it would be displayed in Microsoft Excel.) Follow the link to download a sample SIF file.
• In the first example below, each source node in the network is listed in the first column, the numerical value of the weight is listed in the second column, and the target node is listed in the third column. If a node has multiple target nodes, it is listed multiple times, one relationship per line. Note that FHL1, GTS1, MSN1, and RPH1 do not have any target nodes, but are still listed as source nodes. While they can be listed, it is not necessary to do so, as is shown in the second example.
  
  
  
  
• In the second example below, only source nodes that have target nodes are listed in the first column, the numerical value of the weight is listed in the second column, and the target node is listed in the third column. If a node has multiple target nodes, it is listed multiple times, one relationship per line. (FHL1, GTS1, MSN1, and RPH1 were removed because they did not have any target nodes.)
  
  
  
  
• For a weighted network, each source-to-target relationship should be on a separate line. While it is possible that two or more source-to-target relationships would have the same exact weight value, it is not recommended to use the shorthand of listing multiple target nodes on the same row.
• GRNsight will automatically detect a SIF file representing a weighted network when it finds numerical values as the relationship type. It will change the arrowhead type and the thickness and color of the edges as described in Section 2a.
• However, if it detects a mix of numerical and "pd" string values as the relationship type, it will issue a warning to the user that this was detected, but still display the network as an unweighted network.
• If the user has selected the menu item Edit > Preferences > Default to black edges with regular arrowheads, GRNsight will ignore the numerical weight values and display the network as an unweighted graph.
• Note that these “weighted” SIF files are able to be imported in to Cytoscape (tested with v3.4.0), but the weight values will not be properly stored as edge attributes by Cytoscape because it does not use the SIF format in this way.
• Note that GRNsight is designed to visualize “small-scale” or “medium-scale” gene regulatory networks, not the entire gene regulatory network for an organism. Currently, it is recommended that you import networks with no more than 35 unique genes or 70 edges. If you upload a network with 50 nodes or 100 edges or more, you will receive a warning and the network graph will still display. If you attempt to upload a network of 75 or more nodes or 100 or more edges, the graph will not display.

According to http://graphml.graphdrawing.org, "GraphML is a comprehensive and easy-to-use file format for graphs. It consists of a language core to describe the structural properties of a graph and a flexible extension mechanism to add application-specific data."

Users who have data in GraphML format can import these files directly into GRNsight. However, if you are creating a new input file for use with GRNsight, we recommend using the native Excel format expected by GRNsight as described in Section 1a.

A primer for the GraphML format can be found at http://graphml.graphdrawing.org/primer/graphml-primer.html, whereas the detailed specification can be found at http://graphml.graphdrawing.org/specification.html. The instructions below borrow heavily from the GraphML primer.

• A GraphML file is an Extensible Markup Language (XML) file with the extension .graphml.
• GraphML has the ability to specify graph features that GRNsight cannot display. In those cases, GRNsight will ignore the features that it does not support.
• The minimum requirements for an unweighted GraphML file that GRNsight can read and display is best illustrated by an example. The screenshots below show the top and bottom of the Demo #3 Unweighted GRN (21 genes, 31 edges) GraphML file as exported by GRNsight. (The screenshots were created from a view of this file in the firstobject XML Editor). Click the link to download the file.
  
  
  
  ...
  
  
  
  
• “The first line of the document is an XML process instruction which defines that the document adheres to the XML 1.0 standard and that the encoding of the document is UTF-8, the standard encoding for XML documents.”—GraphML Primer
• “The second line contains the root-element element of a GraphML document: the graphml element. The graphml element, like all other GraphML elements, belongs to the namespace http://graphml.graphdrawing.org/xmlns. For this reason we define this namespace as the default namespace in the document by adding the XML Attribute xmlns="http://graphml.graphdrawing.org/xmlns" to it. The two other XML Attributes are needed to specify the XML Schema for this document. In our example we use the standard schema for GraphML documents located on the graphdrawing.org server. The first attribute, xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance", defines xsi as the XML Schema namespace. The second attribute, xsi:schemaLocation="http://graphml.graphdrawing.org/xmlns http://graphml.graphdrawing.org/xmlns/1.0/graphml.xsd", defines the XML Schema location for all elements in the GraphML namespace.”—GraphML Primer
• The third line is a comment injected by GRNsight, giving the name of the program that created the document (GRNsight), the version (in this case v1.18), URL of the GRNsight home page, and URL to the GitHub release page for the specified version. This information is provided upon export of a GraphML file by GRNsight to assist with keeping track of the provenance of the data, and is optional for the import of GraphML files into GRNsight.
• The next three lines are key elements. GraphML standardizes only the representation of nodes and edges and their directions; all other characteristics, such as names, weights, and other values, are left for others to specify through a key element, which is not subject to a controlled vocabulary. For an unweighted network, GRNsight-exported GraphML defines the keys:
• name for the node label (The node name is defined for compatibility with Cytoscape, but is optional for GRNsight to read and display the graph.)
• interaction for the relationship type of the edge (The relationship type is defined in the same way as for SIF files described in Section 1c for compatibility with Cytoscape, but is optional for GRNsight to read and display the graph.)
• name for the edge (The edge name is defined for compatibility with Cytoscape, but is optional for GRNsight to read and display the graph.)
• "A graph is, not surprisingly, denoted by a graph element. Nested inside a graph element are the declarations of nodes and edges. A node is declared with a node element, and an edge with an edge element."—GraphML Primer
• In the graph element in line 7, the edgedefault attribute is set to “directed”. GRNsight will only display directed graphs, not undirected or mixed graphs.
• The id attribute gives the name of the GraphML file, and is optional for GRNsight to read and display the graph. The id is provided for compatibility with Cytoscape.
• The next three elements in the screenshot are node declarations, giving an id and name for the nodes “ABF1”, “ACE2”, and “AFT1”.
• Note that for GRNsight-exported GraphML, the node id and name are identical and unique to the file. However, other programs may have a node name that is different than the node id. For example, both Cytoscape v3.4.0 and yED v3.16 assign internal node identifiers that the user cannot edit. The user does have the ability edit the node label, and it is for this reason that GRNsight will preferentially display the node label on the graph. However, user editability in Cytoscape and yED leads to a situation where a user can actually assign the same label to two different nodes. In this case, duplicate node labels will be displayed by GRNsight, and the user is directed to go back and edit the source file.
• Cytoscape uses the name key to assign a label to nodes and a shared name key to assign a label to the root network when subnetworks are present.
• yED uses a y:NodeLabel key to assign the label to nodes.
• In order to maximize the ability for GRNsight to read GraphML exported from Cytoscape and yED, the following conditions are followed when displaying a graph: if all three keys are present name gets top priority, followed by shared name, then y:NodeLabel, then id.
• The next three elements shown in the screenshot are edge declarations for the edges ROX1-to-YAP6, YAP1-to-YAP6, and YAP6-to-YAP6.
• The edge element attributes source and target give the node id for the source and target node, respectively.
• The interaction key is the relationship type, similar to what is defined for SIF files. It is given for compatibility with Cytoscape and defaults to “pd” for protein->DNA, but is otherwise ignored by GRNsight for display.
• The edge name key is given for compatibility with Cytoscape, but is otherwise ignored by GRNsight for display.
• As noted above, GraphML has the ability to specify graph features that GRNsight cannot display. In those cases, GRNsight will ignore these features. The original imported file will not be changed, but a GRNsight re-export of the same graph will omit these elements.
• GRNsight cannot display a nested graph. Any links that connect nodes from the main graph to nodes in the subgraph will not display.
• Any other specifications for the display of the graph (layout, colors, etc.) will be ignored by GRNsight.
• We have tested the display of networks exported in GraphML format from Cytoscape v3.4.0 and yED v3.16, but because the flexibility of the GraphML standard enables divergence in implementation, we cannot guarantee that GraphML exported from other programs will be read and displayed correctly by GRNsight. If you encounter problems, please submit an issue to GRNsight @ GitHub (requires free GitHub account).
• The minimum requirements for a weighted GraphML file that GRNsight can read is best illustrated by an example. The screenshots below show the top and bottom of the Demo #4 Weighted GRN (21 genes, 31 edges, Schade et al. 2004 data) GraphML file as exported by GRNsight. Click the link to download the file.
  
  
  
  ...
  
  
  
  
• Note that there are only two major differences between the unweighted and weighted GraphML formats that GRNsight exports:
• A key declaration defining a weight attribute for an edge as the datatype “double”.
• Note that there is currently a bug in Cytoscape v3.4.0 with the way it exports GraphML files. It currently assigns the weight attribute for a node instead of an edge. To maximize the ability for GRNsight to read and display Cytoscape-exported GraphML, GRNsight does not currently enforce that the weight attribute be declared "for an edge". The bug has been reported to the Cytoscape group.
• The addition of the weight key for each edge. For example, the weight for the edge ROX1-to-YAP6 is -0.7503313690638223.
• GRNsight automatically detects if a GraphML file specifying a weighted graph has been imported.
• For a weighted graph to display, the weight attribute must be present and valid for all edges declared in the graph. If one or more weights are missing, GRNsight will display the graph as unweighted with black edges and pointed arrowheads.
• Note that GRNsight is designed to visualize “small-scale” or “medium-scale” gene regulatory networks, not the entire gene regulatory network for an organism. Currently, it is recommended that you import networks with no more than 35 unique genes or 70 edges. If you upload a network with 50 nodes or 100 edges or more, you will receive a warning and the network graph will still display. If you attempt to upload a network of 75 or more nodes or 100 or more edges, the graph will not display.

• Once you have your appropriately formatted Excel Workbook file (*.xlsx), you can load it into GRNsight.
• Go to the GRNsight home page. Select the menu option File > Open. A dialog box will open for you to select your file. Clicking the Open button will load your file. Clicking the Cancel button will return you to the GRNsight home page.
• The name of the file that is currently loaded will appear in the menu bar. The number of nodes and edges in the graph will display on the right-hand side of the menu bar.
• To load a different file, select the menu option File > Open to load a GRNsight-formatted .xlsx file or File > Import to import a file in SIF or GraphML.
• To reload the same file, select the menu option File > Reload. Note that the Reload menu item will be disabled until a file has been loaded.
• If you do not have your own input spreadsheet, you can load and view demonstration gene regulatory networks. Selecting the menu item Demo > Demo #1: Unweighted GRN (15 genes, 28 edges, Dahlquist Lab unpublished data) or Demo #3: Unweighted GRN (21 genes, 31 edges) will load a network graph with black edges and regular (pointy) arrowheads. Selecting the menu item Demo > Demo #2: Weighted GRN (15 genes, 28 edges, Dahlquist Lab unpublished data) or Demo #4: Weighted GRN (21 genes, 31 edges, Schade et al. 2004 data) will load a network graph with colored edges of varying thicknesses. Demos #3 and 4 are based on modeling reported in the following publication: Dahlquist, K.D., Fitzpatrick, B.G., Camacho, E.T., Entzminger, S.D., and Wanner, N.C. (2015) Parameter Estimation for Gene Regulatory Networks from Microarray Data: Cold Shock Response in Saccharomyces cerevisiae. Bulletin of Mathematical Biology. 77: 1457-1492, published online 29 September 2015, DOI: 10.1007/s11538-015-0092-6, which, in turn, is based on microarray data from this publication: Schade B, Jansen G, Whiteway M, Entian KD, Thomas DY (2004) Cold adaptation in budding yeast. Mol Biol Cell 15:5492–5502. doi:10.1091/mbc.E04-03-0167. See Section 2 for details on how GRNsight displays the graph.

• Once you have your appropriately formatted SIF file (*.sif; see Section 1a-ii.c), you can import it into GRNsight.
• Go to the GRNsight beta page. Select the menu option File > Import > SIF…. A dialog box will open for you to select your file. Clicking the Open button will load your file. Clicking the Cancel button will return you to the GRNsight home page.
• The name of the file that is currently loaded will appear in the menu bar. The number of nodes and edges in the graph will display on the right-hand side of the menu bar.
• To reload the same file, select the menu option File > Reload. Note that the Reload menu item will be disabled until a file has been imported.
• To load a different file, select the menu option File > Open to load a GRNsight-formatted .xlsx file or File > Import to import a file in SIF or GraphML format.
• GRNsight will automatically detect a SIF file representing a weighted network when it finds numerical values as the relationship type. It will change the arrowhead type and the thickness and color of the edges as described in Section 2a.
• However, if GRNsight detects a mix of numerical and text string values or missing values for the relationship type, it will display the network as an unweighted network.
• If the user has selected the menu item Edit > Preferences > Default to black edges with regular arrowheads, GRNsight will ignore the numerical weight values and display the network as an unweighted graph.

• Once you have your appropriately formatted GraphML file (*.graphml), you can import it into GRNsight.
• Go to the GRNsight beta page. Select the menu option File > Import > GraphML…. A dialog box will open for you to select your file. Clicking the Open button will load your file. Clicking the Cancel button will return you to the GRNsight home page.
• The name of the file that is currently loaded will appear in the menu bar. The number of nodes and edges in the graph will display on the right-hand side of the menu bar.
• To reload the same file, select the menu option File > Reload. Note that the Reload menu item will be disabled until a file has been imported.
• To load a different file, select the menu option File > Open to load a GRNsight-formatted .xlsx file or File > Import to import a file in SIF or GraphML format.
• GRNsight will automatically detect a GraphML file representing a weighted network when it finds numerical values stored as weight attributes for the edges. It will change the arrowhead type and the thickness and color of the edges as described in Section 2a.
• However, if GRNsight detects a mix of numerical and text string values or missing values for the weight attributes, it will display the network as an unweighted network.
• If the user has selected the menu item Edit > Preferences > Default to black edges with regular arrowheads, GRNsight will ignore the numerical weight values and display the network as an unweighted graph.