Related Experiment Video
Updated: Nov 7, 2025
11:19
Label-Free Immunoprecipitation Mass Spectrometry Workflow for Large-scale Nuclear Interactome Profiling
Published on: November 17, 2019
16.6K
PubMed-Scale Chemical Concept Embeddings Reconstruct Physical Protein Interaction Networks
Blaž Škrlj1,2, Enja Kokalj1,2, Nada Lavrač2,3
1Jožef Stefan International Postgraduate School, Ljubljana, Slovenia.
Frontiers in Research Metrics and Analytics
|April 30, 2021
Summary
A new network, CHEMMESHNET, uses chemical MeSH annotations from PubMed to predict protein-protein interactions. This literature-based approach successfully reconstructs known interactions and identifies novel ones.
Area of Science:
- Biomedical Informatics
- Computational Biology
- Network Science
Background:
- PubMed contains over 25 million biomedical documents, making it challenging for experts to track all relevant literature.
- Knowledge gaps arise from the inability to keep up with the vast amount of novel scientific publications.
- Existing protein-protein interaction (PPI) networks often lack comprehensive coverage or are limited by experimental data.
Purpose of the Study:
- To develop CHEMMESHNET, a novel PubMed-based network of chemical-protein associations.
- To demonstrate that latent representations learned from literature data can reconstruct known physical protein-protein interactions.
- To leverage the network for prioritizing and identifying potentially novel protein-protein interactions.
Main Methods:
- Construction of CHEMMESHNET using expert-curated MeSH annotations of chemicals from all available PubMed articles, resulting in over 10 million associations.
- Learning latent representations of concepts within the CHEMMESHNET network.
- Utilizing simple linear embeddings of node pairs coupled with a neural network classifier to reconstruct known protein-protein interactions.
- Prioritizing novel interactions based on common chemical context derived from learned representations.
Main Results:
- CHEMMESHNET was constructed with over 10 million chemical-protein associations.
- Latent representations learned from literature data were sufficient to reconstruct a significant portion of known empirically determined protein-protein interactions.
- A machine learning model reliably reconstructed existing protein-protein interactions.
- The method successfully prioritized novel protein-protein interactions, with top-ranked interactions showing potential for complex formation and aligning with structure-based predictions.
Conclusions:
- Literature-based network representations, specifically CHEMMESHNET, are effective for reconstructing known protein-protein interactions.
- The developed approach provides a powerful tool for discovering novel, biologically relevant protein-protein interactions by analyzing chemical contexts.
- CHEMMESHNET offers a scalable solution to bridge knowledge gaps in the rapidly expanding biomedical literature.
Keywords:
PubMeddata-miningknowledge graphsliterature-based discoverymachine-learningrepresentation learningMore Related Videos
08:38Genome-wide Protein-protein Interaction Screening by Protein-fragment Complementation Assay PCA in Living Cells
Published on: March 3, 2015
13.6K
07:08Optimization of Synthetic Proteins: Identification of Interpositional Dependencies Indicating Structurally and/or Functionally Linked Residues
Published on: July 14, 2015
7.4K
Related Concept Videos
02:04Protein-protein Interfaces
14.1K
Many proteins form complexes to carry out their functions, making protein-protein interactions (PPIs) essential for an organism's survival. Most PPIs are stabilized by numerous weak noncovalent chemical forces. The physical shape of the interfaces determines the way two proteins interact. Many globular proteins have closely-matching shapes on their surfaces, which form a large number of weak bonds. Additionally, many PPIs occur between two helices or between a surface cleft and a...
14.1K
02:04Protein-Protein Interfaces
4.1K
4.1K
02:26Protein Networks
4.2K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.2K
02:26Protein Networks
2.6K
2.6K
01:24Protein Organization
8.2K
Proteins are polymers of amino acid residues. They are versatile and responsible for different cellular functions, including DNA replication, molecular transport, catalysis, and structural support. Proteins have a hierarchical structure comprising at least three levels of organization: primary, secondary, and tertiary structure. Some large proteins have a quaternary structure where individual protein subunits are linked together.
The primary structure of a protein is its amino acid sequence....
The primary structure of a protein is its amino acid sequence....
8.2K
01:49Conserved Binding Sites
4.7K
Many proteins’ biological role depends on their interactions with their ligands, small molecules that bind to specific locations on the protein known as ligand-binding sites. Ligand-binding sites are often conserved among homologous proteins as these sites are critical for protein function.
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
4.7K