Protein Kinase Ontology

Current version: 67
Date: 2023-10-11

Protein Kinase Ontology (ProKinO) is a collaborative effort between the Evolutionary Systems Biology Group Lab of Dr. Natarajan Kannan at the Biochemistry and Molecular Biology Department and Dr. Krys J. Kochut's lab at the School of Computing, both at the University of Georgia, Athens, USA. Dr. Gurinder Gosal, formerly a student in Computer Science, UGA, created an initial version of the software system to automatically populate ProKinO from the selected data sources (see Data sources, below). The population system has been subsequently enhanced by our students Dr. Shima Dastgheib, formerly Computer Science, UGA, Dr. Daniel McSkimming, formerly IOB, UGA, and Dr. Abbas Keshavarzi, formerly Computer Science, UGA, and several other current and former lab members (see People, below).

The conceptual representation of such diverse information in one place, in the form of the ProKinO knowledge graph, enables not only rapid discovery of significant information related to a specific protein kinase, but also enables large scale integrative analysis of the protein kinase family. An outline of the data organization in ProKinO is represented using a UML class diagram (see Schema, below).


Soleymani, S., Gravel, N., Huang, L. C., Yeung, W., Bozorgi, E., Bendzunas, N. G., Kochut, K.J., and Kannan, N. (2022). Dark kinase annotation, mining and visualization using the Protein Kinase Ontology. bioRxiv, 2022-02. doi:

Salcedo, M., Gravel, N., Keshavarzi, A., Huang, L., Kochut, K., Kannan, N. (2023). Predicting Protein and Pathway Associations for Understudied Dark Kinases using Pattern-constrained Knowledge Graph Embedding. PeerJ, under review

Yeung, W., Zhou, Z., Li, S., & Kannan, N. (2023). Alignment-free estimation of sequence conservation for identifying functional sites using protein sequence embeddings. Briefings in Bioinformatics, 24(1), bbac599.

McSkimming, D.I., Dastgheib, S., Talevich, E., Narayanan, A., Katiyar, S., Taylor, S.S., Kochut, K.J., Kannan, N. (2015). "ProKinO: a unified resource for mining the cancer kinome." Human Mutation 36(2):175-86. doi:10.1002/humu.22726.

Gosal, G., Kochut, K.J., Kannan, N. (2011). "ProKinO: An Ontology for Integrative Analysis of Protein Kinases in Cancer." PLoS ONE 6(12):e28782. doi:10.1371/journal.pone.0028782

Gosal, G., Kannan, N., Kochut, K.J. (2011). "ProKinO: A Framework for Protein Kinase Ontology," in Proceedings of the 2011 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 550-555.

Data Sources

The following table shows the versions of the datasources used for population of the ProKinO version loaded in this browser.

KinBaseAugust, 2012
PseudokinasesSeptember, 2018
MSADecember, 2015
Ligands dataSeptember, 2021
Sub-domainsJune, 2023

Disease Data COSMIC (Catalogue of Somatic Mutations in Cancer) database ` is used as the source of information regarding kinase cancer mutations. In addition to mutations, other information such as, the primary sites, histology, samples, description and other relevant features is also obtained from COSMIC.

The mutation data was obtained from the Sanger Institute Catalogue Of Somatic Mutations In Cancer web site, Bamford et al (2004) The COSMIC (Catalogue of Somatic Mutations in Cancer) database and website. Br J Cancer, 91,355-358.

Pathway and Reaction Data The pathway and reaction data in ProKinO is obtained from Reactome.

Reactome project. (26th January 2011).

Function Data The information regarding functional domains and functional features associated with kinase domains is obtained from UniProt. The information regarding the functional domains associated with kinase domains, the crystal structures solved for each kinase domain, isoforms, and functional features associated with kinases, such as the modified residue, signal peptide, topological domain, cellular location and tissue specificity, are obtained from UniProt.

The UniProt Consortium, Ongoing and future developments at the Universal Protein Resource Nucleic Acids Res. 39: D214-D219 (2011).

Classification and Sequence Data Data regarding protein kinase sequence and classification is obtained from KinBase.

The Protein Kinase Complement of the Human Genome, G Manning, DB Whyte, R Martinez, T Hunter, S Sudarsanam (2002). Science 298:1912-1934.

Pseudokinase data and their classification Pseudokinase sequences were identified from the Uniprot reference proteomes database and classified into pseudokinase families in the Kannan Lab at the University of Georgia. Pseudokinase protein names were either obtained from UniProt or assigned by the pseudokinase classification if the protein name is not provided by UniProt.

Kinase Sub-domain Data To capture the sub-domain information in ProKinO, we have used a motif model [1, 2] with key motifs corresponding to each of the sub-domains in the kinase domain [3].
  1. Neuwald AF, Liu JS, Lawrence CE. Gibbs motif sampling: detection of bacterial outer membrane protein repeats. Protein Sci. 1995 Aug;4(8):1618-32. PubMed PMID:8520488; PubMed Central PMCID: PMC2143180.
  2. Kannan N, Neuwald AF. Did protein kinase regulatory mechanisms evolve through elaboration of a simple structural component? J Mol Biol. 2005 Sep 2;351(5):956-72. PubMed PMID: 16051269.
  3. Hanks SK, Hunter T. Protein kinases 6. The eukaryotic protein kinase superfamily: kinase (catalytic) domain structure and classification. FASEB J. 1995 May;9(8):576-96. Review. PubMed PMID:7768349.

ProKinO Schema

The picture below is an outline of the ProKinO schema as a UML class diagram. It shows major ProKinO classes and object properties (relationships connecting ProKinO classes). It is available in PDF, as well

ProKinO Schema

ProKinO Browser

You can use this ontology browser to quickly locate protein kinase proteins and a lot of information related to the proteins, including the sequence, structure, function, mutation and pathway information on kinases.

You may initiate ProKinO browsing by selecting one of the following items in the Browse menu:

  • Organisms, which provides the list of organisms; select an organism of interest and click on it to display the kinase proteins for the organism. Subsequently, you can select and click one of the proteins to explore.
  • Proteins, which provides the list of kinase proteins for many organisms; select a protein of interest and click on it to display the protein data.
  • Kinase Domains, which provides the hierarchy all kinase domains; you will be able to expand the groups into families and subfamilies, eventually finding the domains; again, select one of interest and click on it to display the kinase domain information.
  • Diseases, which provides the list of all diseases (limited to cancers at this time) that are affected by kinase proteins; select one of interest and click on it to display the cancer data.
  • Functional Domains, which provides the list of all functional domains; again, you may select one of interest and click on it to display the functional domain data.

It is also possible to initiate ProKinO browsing by searching for a specific kinase protein, disease, pathway, or for any object. The searching uses a set of terms, possibly separated by OR, AND and AND NOT boolean connectors. It is possible to use wildcards, for example, FGFR*. Phrases should be enclosed within double quotes.

The Query functionality for Proteins, Arbitrary SPARQL queries can be submitted (however, the size of the resulting set is limited to 1500, at this time).

The ProKinO ontology has not been released for downloading, yet.

After you initiate ProKinO browsing by selecting and displaying the data information page for one of the proteins, kinase domains, diseases, or functional domains, you will be able to explore the ontology by activating the hyper-links leading to a variety of related data.

Usually, a data information page includes the properties of the shown entity (for example, cellular location or tissue specificity of the proteins), and links to other related ontology entities (for example, functional domains, sequences, pathways of the proteins).

The protein data is subdivided into sub-areas, including the general information, functional features, mutations, pathways, and external references. These sub-areas of information are available by clicking on the provided tabs. The mutations are further subdivided into substitutions, insertions, deletions, complex mutations, and other.


ProKinO is a collaborative effort between the Evolutionary Systems Biology Group Lab of Dr. Natarajan Kannan at the Biochemistry and Molecular Biology Department and Dr. Krys J. Kochut's lab at the Computer Science Department, both at University of Georgia, Athens, USA. Gurinder Gosal, also at UGA, created the initial version of the software system to automatically populate ProKinO from the selected data sources.

Current Team Members
Natarajan Kannan
Dr. Natarajan Kannan
Krzysztof Kochut
Dr. Krzysztof Kochut
Computer Science
Saber Soleymani
Dr. Saber Soleymani (CS)
Visiting Assistant Prof.
Georgia State University
Amitabh Priyadarshi
Nathan Gravel
Research Assistant
George Bendzunas
George Bendzunas
Postdoctoral Scholar
Amitabh Priyadarshi
Elika Bozorgi
Teaching Assistant
Computer Science
Amitabh Priyadarshi
Amitabh Priyadarshi
Research Assistant
Computer Science
Safal Shrestha
Safal Shrestha
Research Assistant
Former Team Members
Liang-Chin Huang
Dr. Liang-Chin Huang (IOB)
Senior Data Scientist
Melax Tech
Abbas Keshavarzi
Dr. Abbas Keshavarzi (CS)
ML Software Engineer
Annie Kwon
Dr. Annie Kwon (IoB)
Postdoctoral Scholar
Daniel McSkimming
Dr. Daniel McSkimming (IoB)
Director, Bioinf. &
Comp. Biology, USF
Shima Dastgheib
Dr. Shima Dastgheib (CS)
Technical Lead
Gurinder Gosal
Dr. Gurinder Gosal (CS)
Researcher, Univ. of
British Columbia
Reshmi De
Reshmi De (CS)
Senior Software Engineer
Morgan Stanley
Anish Narayanan
Anish Narayanan, M.D. (IoB)
UTHSC-San Antonio
Bhargabi Chakrabarti
Bhargabi Chakrabarti (CS)
Ph.D. Student, ECE
George Mason Univ.