A Semantic Design for the Biological Processes
Associated with Intrinsically Disordered Proteins
Atsuko Yamaguchi1,∗,† , Yumiko Kado2 , Shigetaka Sakamoto3 , Satoshi Fukuchi4 and
Motonori Ota2
1
  Tokyo City University, Tokyo 158-8557, Japan
2
  Nagoya University, Nagoya 464-8601, Japan
3
  HOLONICS Corporation, Numazu 411-0803, Japan
4
  Maebashi Institute of Technology, Maebashi 371-0816, Japan


                                         Abstract
                                         Intrinsically disordered proteins (IDPs) challenge the traditional view of protein structure, lacking a
                                         stable tertiary structure yet playing pivotal roles in various biological processes. To address the growing
                                         need for a comprehensive resource, we developed IDEAL (https://ideal-db.org/), a curated database of
                                         IDPs. IDEAL leverages semantic web technologies like RDF and SPARQL to enhance accessibility and
                                         interoperability. Here, we introduce the design of the newly constructed RDF dataset that represents the
                                         biological processes associated with IDPs.

                                         Keywords
                                         Semantic Web, Resource Description Framework, Intrinsically disordered proteins


1. Introduction
Intrinsically disordered proteins (IDPs) form a dynamic protein class lacking a fixed three-
dimensional structure in the isolated state, allowing diverse conformations upon interact-
ing with other molecules. Vital in cellular processes like signaling and regulation, IDEAL
(http://www.ideal-db.org/) is a significant repository for experimentally verified IDPs and intrin-
sically disordered regions (IDRs) [1]. Our contribution to IDEAL includes a semantic web-based
resource with RDF and SPARQL technologies, providing both a user-friendly website and down-
loadable XML data. Recognizing IDPs’ growing importance, we have expanded our RDF dataset
to cover associated processes. Leveraging knowledge graphs for dynamic and procedural repre-
sentation, this poster paper introduces our knowledge graph design for biological processes
involving IDPs.

2. Representation of IDP-Associated Processes
In order to depict processes associated with IDPs, we have introduced two primary classes:
”State” and ”BiologicalProcess.” The ”State” class encompasses components such as proteins,

SWAT4HCLS 2024
∗
    Corresponding author.
Envelope-Open atsuko@tcu.ac.jp (A. Yamaguchi); ykado@force.cs.is.nagoya-u.ac.jp (Y. Kado); sakamoto@holonics.jp
(S. Sakamoto); sfukuchi@maebashi-it.ac.jp (S. Fukuchi); mota@i.nagoya-u.ac.jp (M. Ota)
Orcid 0000-0001-7538-5337 (A. Yamaguchi)
                                       © 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
    CEUR
    Workshop
    Proceedings
                  http://ceur-ws.org
                  ISSN 1613-0073
                                       CEUR Workshop Proceedings (CEUR-WS.org)
nucleic acids, etc., including their complexes. In the case of a component being a complex, the
subunits are categorized by the ”Subcomponent” class. The ”BiologicalProcess” class is linked to
”ProcessType,” with instances including ”catalyze,” ”associate,” ”dissociate,” ”translocate,” among
others. Furthermore, the ”BiologicalProcess” class is associated with the ”Object” class, which
identifies the facilitator of the given biological process.
   Figure 1 illustrates two instances of these processes. The left panel demonstrates a process
where a complex is formed from the IDP APP and APBB1. APP and APBB1 are classified by the
”Component” class in the first ”State,” connected to a ”BiologicalProcess” with ”ProcessType” as
”associate.” This ”BiologicalProcess” is linked to the second ”State,” which contains a complex of
APP and APBB1 as its components. Similarly, the catalyze process of phosphorylation facilitated
by MAPK10 can be represented using the ”State” and ”BiologicalProcess” classes.




Figure 1: Instances of biological processes associated with an IDP.


3. ProcessType Verification
In addition to relying on the notes of ProcessType in XML and RDF, we can verify it by examining
the relationship between the two connected ”States” and the corresponding ”BiologicalProcess.”
For instance, if the first ”State” comprises two protein components, and the second ”State”
includes a complex, the inferred ”ProcessType” would be ”association.” In cases where the
first ”State” involves a protein and the second ”State” involves a phosphorylated protein, the
determined ”ProcessType” would be ”catalyze.” We have designed such rules for verifying the
”ProcessType” based on the characteristics of the associated ”States.”

4. Acknowledgments
This work was supported by JSPS KAKENHI grant number 21K12148 and 20H05932..

References
[1] S. Fukuchi, T. Amemiya, S. Sakamoto, Y. Nobe, K. Hosoda, Y. Kado, S. D. Murakami, R. Koike,
    H. Hiroaki, M. Ota, IDEAL in 2014 illustrates interaction networks composed of intrinsically
    disordered proteins and their binding partners, Nucleic Acids Research 42 (2014).