KNeXT: a NetworkX-based topologically relevant KEGG parser.

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Recreating gene and mixed gene-compound networks from Kyoto Encyclopedia of Genes and Genomes (KEGG) Markup Language (KGML) files presents challenges.
• Current data structures obscure the original topological environment, leading to inaccurate network representations.
• Overlapping accession numbers and duplicated identifiers cause artificial collapse of network neighborhoods.

Purpose of the Study:

• To develop a novel Python-based parser, KNeXT, for accurate recapitulation of genetic and mixed networks from KGML data.
• To overcome the limitations of existing parsers in preserving the topological environment of biological networks.
• To ensure high-fidelity topological representations and programmatic access to original node positions.

Main Methods:

• Developed a Python-based KEGG NetworkX Topological (KNeXT) parser.
• Ingested KGML files via built-in APIs to dynamically create topological representations.
• Utilized NetworkX framework to generate tab-separated files compatible with other graph frameworks.
• Enabled parsing of local folders or single files and conversion to NCBI or UniProt IDs.

Main Results:

• KNeXT accurately recapitulates genetic and mixed networks from KGML map data.
• The parser preserves the original topological environment and node positions (x-y axis).
• Generated tab-separated files maintain programmatic access to pathway data and are importable into other graph frameworks.
• KNeXT supports parsing of local files/folders and ID conversion (NCBI/UniProt).

Conclusions:

• KNeXT offers a robust solution for accurately reconstructing KEGG pathways, overcoming limitations of previous methods.
• The tool ensures the preservation of original network context and topological details.
• KNeXT enhances the reliability of pathway visualization and analysis in bioinformatics research.