View source for OntologySummit2017 Communique/Hierarchical Systems

== Functionalization of Hierarchical Systems for making Inferences ==

=== Opportunities ===

The cell, along with body systems is usually modeled as a hierarchical system and can be represented using ontologies for the different levels of the biological hierarchy -from the molecular and biochemical level to the cellular and tissue level to the organ and organ system level.   By functionalizing a hierarchical system one can reason by means of biological inferences, such as the the genotype-phenotype map.  ML guided by ontologies, such as the manually cured Gene Ontology (GO), is the underlying technique employing extensive measurement data on the network relation between gene and to their protein products. GO is uniquely structured to handle this biological issue because it is organized to the natural categories on Biological Process (BP), Cellular Component (CC), Molecular Function (MF).
The technique leverages knowledge of gene co-expressions and protein-protein interactions to create a data-derived gene similarity network. An alignment process identifies which data terms are  and which and which recapitulate existing knowledge in GO.  Taken together this knowledge can be reduced to a hierarchy and aligned with the Gene Ontology suggesting names for new, data-driven terms. A majority, about 60%, of relations found by data derivation for cellular components are already in GO-CC, but only about 25% of these derived terms for Biological Process and Molecular Function were already found in GO, meaning that much potentially useful knowledge was uncovered.

=== Challenges ===
Not every domain has the degree of agreement on base ontologies that has been agree in the BioMed domain.  Likewise not every area has the degree of hierarchy.  Therefore it is an open research issue whether something like this be achieved in other domains.

=== Future Prospects === 
Among the interesting aspects of this work is that it illustrates not only the extraction of new knowledge from ML and reasoning, but also the value of leveraging existing quality knowledge such as in GO as part of this process and alignment with new knowledge with existing ontologies.
A good prospect is the use of upper domain reference ontologies as starting points for quality ontologies. Possible areas for this work include with Smart City Indicators and Hydro ontologies.

===References ===

[http://ontologforum.org/index.php/DomainVocabularies orsten Hahmann (University of Maine) "Domain Reference Ontologies vs. Domain Ontologies: What's the Difference? Lessons from the Water Domain"]