LipidSig 2.0: Integrating Lipid Characteristic Insights into Advanced Lipidomics Data Analysis

Network analysis


In the network analysis section, five distinct types of networks are provided.
Results from 'Differential Expression' are further examined with the 'Pathway Activity Network', which computes flux changes in the lipid reaction network and helps identify active or suppressed pathways. The 'Lipid Reaction Network' graphically represents significant lipid classes or species within lipid biosynthetic pathways. Lastly, the 'GATOM Network' isolates significant subnetworks within the constructed metabolite-level network.
Note: The 'Pathway activity network', 'Lipid reaction network', and 'GATOM network' are currently available only for two-group comparisons. Multi-group analysis support for these networks is under development.
In the investigated lipid classes, the 'Reactome network' aggregates reactions pertinent to these classes as documented in Reactome. The 'Lipid-related gene enrichment' analysis identifies genes associated with the lipid classes of interest, followed by an enrichment analysis to discern significant pathways.

Demo dataset source: A nutritional memory effect counteracts benefits of dietary restriction in old mice (Nat Metab. 2019)

Pathway activity network

The Pathway activity network module implements a flux‐based pathway analysis algorithm to quantify activity changes across lipid metabolic pathways. In this visualization, each node represents a lipid class and each directed edge a biochemical reaction; edge color encodes the computed pathway activity score (red for up-regulated/positive, blue for down-regulated/negative).

An accompanying table details every reaction step, its activity score, and the genes involved.

It is critical to acknowledge that lipid biochemical pathways exhibit significant variability across mammalian species, contrary to common assumptions. Thus, caution should be used when generalizing findings across species, recognizing that each may present unique lipidomic landscapes that could impact the interpretation and applicability of research outcomes.

  • The "Pathway activity network" module is currently available only for two-group comparisons. Support for multi-group analysis is under development.
  • For input data preparation, please refer to FAQ. You can download inputs directly from the Differential expression analysis to ensure format accuracy.
  • Manipulate the network by clicking on the toolbar at the bottom right and left.
  • View the corresponding information of a lipid or reaction by hovering the mouse on a specific node or line.





Lipid reaction network

The Lipid reaction network renders differential lipid metabolism as a reaction-centric graph in which square nodes represent lipid classes (sized by significance: larger squares for smaller p-values) and circular nodes represent individual lipid species. Yellow squares and red circles denote up-regulated classes and species, respectively; purple squares and blue circles denote down-regulated ones; grey indicates non-significant or missing data.

Two distinct edge types appear: bold orange arrows trace actual biochemical reactions from one lipid entity to another, while thin grey lines simply link each class to its member species.

Beneath the graph are two interactive tables: the edge table details every reaction path with its enzyme EC number and associated genes, and the node table reports log2 fold-change and p-value for each class and species.

  • The "Lipid reaction network" module is currently available only for two-group comparisons. Support for multi-group analysis is under development.
  • For input data preparation, please refer to FAQ. You can download inputs directly from the Differential expression analysis to ensure format accuracy.
  • Manipulate the network by clicking on the toolbar at the bottom right and left.
  • View the corresponding information of a lipid or reaction by hovering the mouse on a specific node or line.




GATOM network

The GATOM network highlights key reactions involving differentially expressed lipid species. These lipids, identified by distinct abundance patterns, are mapped onto a comprehensive metabolic reaction network to construct an experiment-specific framework.

Each node in the network represents a differentially expressed lipid, with color indicating the direction and magnitude of fold change—red for upregulation and blue for downregulation. The size of each node corresponds to its statistical significance, with larger nodes representing smaller p-values. Edges in the network denote biochemical reactions connecting these lipid species.

Below the network, two accompanying tables provide detailed annotation. The edge table summarizes the metabolic paths and reactions connecting lipid species, along with their associated enzymes and genes. The node table lists differentially expressed lipids included in the network, displaying their fold change (Log2FC) and statistical significance (p-value). This integrated framework allows users to interpret lipid metabolic alterations at both the molecular and pathway levels.

  • The "GATOM network" module is currently available only for two-group comparisons. Support for multi-group analysis is under development.
  • For input data preparation, please refer to FAQ. You can download inputs directly from the Differential expression analysis to ensure format accuracy.
  • Manipulate the network by clicking on the toolbar at the bottom right and left.
  • View the corresponding information of a lipid or reaction by hovering the mouse on a specific node or line.



Reactome network

The 'Reactome Network' contains the interaction of multiple lipids and genes and is summarized from the Reactome database, a curated database encompassing pathways and reactions in human biology. This function supports users to discover the metabolic networks between small molecules within Reactome Network topologies and their interactions with proteins. We integrated the reactions in these pathways and explored the inputted lipids' metabolic flow and regulatory proteins. This information will be used to build an interactive network containing different nodes, including biochemical reactions (square), small molecules (diamond), proteins (round), or complexes (triangle).
The table below the network provides the corresponding information of the network, such as path, path rank, node order, and node name.

  • Nodes representing the selected lipid classes are highlighted in larger size and red.
  • Manipulate the network by clicking on the toolbar at the bottom right and left.
  • Display or hide a specific node or line by clicking on it.



Reactome network table



Lipid-related gene enrichment

This function network of the user-defined lipid-related gene sets is presented by summarising the pathways from KEGG and Reactome databases. The mapped genes of the user-selected lipid class will be enriched and illustrated in the circular network diagram, which allows users to explore the significant functions and the relationship between lipid-related pathways and genes based on their lipid classes. The larger node size refers to more gene numbers.
The table below the network provides the corresponding information of the network, such as pathway, lipid gene number, pathway gene number, and -log10 p-value.

  • Manipulate the network by clicking on the toolbar at the bottom right and left.
  • View the corresponding information of a node or line by hovering the mouse on it.



Lipid-related gene enrichment table