ScyNet

ScyNet is a Cytoscape app for visualizing community metabolic models. ScyNet reduces the visualized network to community members and their exchange metabolites.

Features

• ScyNet drastically reduces community network size (>90% for genome scale metabolic models expected)
• The resulting networks allow for better overview and creation of figures
• ScyNet allows for contextualizing the reduced networks with flux data (e.g. from FBA or FVA results)
• The network styling uses a color-blind friendly palette
• The custom layout algorithm provides a good overview of the community members and their exchange metabolites

Requirements

Please make sure that the following requirements are installed

• Cytoscape version 3.9.0 or later
• cy3sbml (for importing metabolic models)

Installation

ScyNet is available in the Cytoscape App Store. It can be installed within Cytoscape by going to Apps -> App Manager and then search for ScyNet. Once found, select ScyNet and click install.

Basic usage

For more detailed information and examples, visit the ScyNet wiki!

Loading a Model

Import the input community metabolic model with the cy3sbml app (see requirements). This can be done by going to File -> Import -> Network from file and then selecting the community metabolic model SBML file. Please be patient when loading models, especially larger models (>10 members). This process can take a couple of minutes, depending on your hardware.

Model Specifications

ScyNet requires some information in the community metabolic model SBML file to be in a specific format. This is required to correctly attribute metabolites to either the community members or the shared exchange compartment. SBML files need to fulfill the following:

1. There must be a shared exchange compartment containing all exchange metabolites and their respective boundary reactions, as well as transfer reactions to the respective member compartments. To detect this compartment there needs to be either
   1. A shared exchange compartment named medium.
   2. A parameter shared_compartment_id which is set to the name of the shared exchange compartment.
2. Each metabolite ID must be prefixed with the identifier of the community member they are associated with followed by an underscore: memberId_metaboliteId
3. Each compartment ID that is part of a community member must be prefixed with the identifier of the community member they are associated with, followed by an underscore: memberId_compartmentId

A metabolic model with this format can be generated from member models using the PyCoMo package.

Creating a Reduced Network

After importing the community metabolic network with cy3sbml, 3 networks should be available: All, Base, and Kinetic. Select either the All or Kinetic network and run the network simplification via Apps -> ScyNet -> Create Simplified Community Network.

Layout and Styling

ScyNet offers several options for changing the network layout, all of which can be found under Apps -> ScyNet. To run them, a network created by ScyNet needs to be selected first.

• Contextualize with Flux Data (see below)
• Apply ScyNet Layout Places all nodes into concentric circles based on node type and connection to community member nodes. This layout is automatically applied when creating a simplified community network.
• Toggle Non-Cross-Fed Metabolite Visibility Hides all metabolite nodes that are not cross-fed. If all non-cross-fed metabolite nodes are hidden, it reveals them instead. Only works if flux data is available.
• Toggle Edge Width Relative to Flux Sets edge widths relative to the corresponding flux values. Running this again will set all edge widths to the default width. Only works if flux data is available.
• Toggle Zero Flux Edge Visibility Hides all edges with a flux value of 0. If all edges with 0 flux are hidden, it reveals them instead. Only works if flux data is available.

Contextualization with flux data

ScyNet can contextualize the edges of the community network with flux data. This can be either single value fluxes (such as from FBA) or flux ranges (such as from FVA). To read the flux values with ScyNet, they need to be supplied as tab separated files (further requirements below).

FBA Flux File

The flux vector of a single state can be visualized by providing the vector in a tab separated file. This file needs to contain two columns, called reaction_id and flux.

FVA Flux File

Also flux ranges can be visualized with ScyNet. For doing so, a tab separated file with three columns needs to be provided: reaction_id, min_flux, max_flux.

Citing ScyNet

Michael Predl, Kilian Gandolf, Michael Hofer, Thomas Rattei, ScyNet: Visualising interactions in community metabolic models, Bioinformatics Advances, 2024;, vbae104, https://doi.org/10.1093/bioadv/vbae104