DancePartner: Python Package to Mine Multiomics Relationship Networks from Literature and Databases

Area of Science:

• Systems Biology
• Bioinformatics
• Computational Biology

Background:

• Multiomics experiments aim to elucidate molecular biology alterations between conditions.
• Understanding biomolecule relationships (e.g., interactions, metabolic pathways) is crucial.
• Existing databases lack comprehensive relationship data, especially for understudied species, necessitating literature mining.

Purpose of the Study:

• To develop an accessible tool for extracting biomolecule relationships from literature and databases.
• To create a Python package, DancePartner, for automated relationship extraction and multiomics network construction.
• To enable mapping of biomolecule synonyms to standardized identifiers and visualization of networks.

Main Methods:

• Developed the DancePartner Python package for literature and database mining.
• Implemented functions for biomolecule synonym mapping and network visualization.
• Utilized large datasets of papers and abstracts for *Caenorhabditis elegans* and *Saccharomyces cerevisiae*.

Main Results:

• Successfully extracted and integrated biomolecule relationships from literature and public databases (KEGG, WikiPathways, UniProt, LipidMaps).
• Demonstrated DancePartner's capability using extensive datasets for *C. elegans* and *S. cerevisiae*.
• Visualized and characterized the resulting multiomics networks, comparing network build times.

Conclusions:

• DancePartner provides an accessible solution for constructing multiomics networks, particularly for species with limited existing data.
• The package facilitates the integration of diverse data sources for comprehensive biological network analysis.
• Automated literature mining and network visualization are key for advancing multiomics research.