Ontology Development Methodologies – Uschold and King

Ontology Development Methodologies take two.

Pictorially, Uschold and King’s methodology can be summarized as follows:

As can be seen, the process can be broken down into a number of discreete steps: identification of the ontology’s purpose, ontology capture, ontology coding, integration of existing ontologies, ontology evaluation and ontology documentation.

Let’s look at each of these steps in turn:

Identification of the Ontology’s Purpose

All methodologies for the development of ontologies that will be surveyed in this series of blog posts agree, that the definition of the purpose is of vital importance as it essentially helps to define scope and granularity of the ontology. However this is where agreements usually ends and while Grueninger and Fox are relatively prescriptive about how this could be achieved, Uschold and King essentially leave the puropose and scope definition up to the ontological engineer. They do, however, discuss a number of purposes which have been reported in the literature and thus provide some criteria which an ontological engineer could consider when attempting to formulate a “mission statement” for his ontology. These are:

• definition of a vocabulary?
• meta-level specification of a logical theory?
• ontology primarily intended for use of a small group?
• re-use by large community?
• ontology a means to structure a kowledge base?
• ontology part of knowledge base?
• …

Ontology Capture

Uschold and King define ontology as a process, which can again be broken down into a number of smaller steps:

• identification of the key concepts and relationships in the domain of interest
• production of precise, unambiguous text definitions for such concepts and relationships
• identification of of terms to refer to such consepts and relationships
• achieving community agreement on all of the above

For the initial stages of the ontology capture process, Uschold and King recommend a brainstorming phase, which should produce all relevant concepts and relationships the ontology should contain. At this stage, concepts are represented by terms (labels) which my hide differences in interpretation and understanding of fundamental concepts. Furthermore, the authors point out that, while, in their experience, brainstorming works well, it may have to be supplemented with other sources of information if domain expertise is required.

In a second step then, the identified concepts should be arranged into “work areas corresponding to naturally arising sub-groups”. To decide whether a term should be included or excluded from a grouping and the ontology in general, a reference should be made to the requirements specification of the ontology. The authors thus underline again the vital importance of the availability of such a document. They furthermore recommend, that inclusion or exclusion decisions be documented for future reference. Finally, it is recommended that “semantic cross-references” be identified which link concepts in one group to those of another group.

In a third step in the capture process, Uschold and King recommend the identification of metaontologies which may be suitable for the particular domain ontology to be constructed, without, at this stage, making a firm ontological commitment. They recommend that a consideration of the concepts in the domain ontology and their interrelationships guide the choice of a metaontology.

In a fourth step, precise definitions of all terms and concepts in the ontology should be produced. For this purpose, the authors recommend that defintions for concepts which have a maximum semantic overlap between work areas should be produced first, as these are more likely to be the most important concepts and it is important to get these definitions right in the first instance. Furthermore, Uschold and King advocate to focus initially on the definition of cognitively basic terms as opposed to more abstract ones, arguing that this should facilitate the process of relating terms in different areas. To develop the definitions, Uschold and King recommend that precise natural language text definitions of all terms be produced, while takinng great care to ensure consistency with other terms which are already in use. Furthermore, the introduction of new terms to to be avoided at this stage. The provision of examples is considered to be helpful.

Possible guidelines for dealing with ambiguous or hard terms are also provided. These are:

• Do not use the term if it is too ambiguous to define and consensus cannot be reached.
• Before attempting a definition, clarify the underlying idea. This can, for example be done by consulting dictionaries and trying to avoid technical terms as much as possible.
• To avoid getting hung up on concept labels, give term meaningless labels such as x1, x2, x3 and then attempt the definition of the underlying idea.
• If a definition/clarification has been achieved, attach a term to the concepts, which, if possible avoids the previous ambiguous term.

Once this has been done, an ontological commitment to a meta-ontology should be made at this stage, which will support the following coding stage.

Ontology Coding

Ontology coding denotes the representation of the conceptualization which has been developed during the capture phase in a formal language. In essence, this involves three sub-stages: a commitment to a meta-ontology, the choice of a formal ontology language and the coding itself. Furthermore, existing ontologies which are to be re-used should be included during the coding stage. In general, the merging of two different ontologies is no trivial task, but can be facilitated, if a meta-ontology is available in all used ontologies to provide some sort of “schema definition”.

Ontology Evaluation

Ontology evaluation is in essence a technical judegement of the performance of an ontology w.r.t. its requirements specification, its ability to answer questions and its associated software environment.

Ontology Documentation

We have already alluded to the fact that decisions concerning the inclusion or exclusion of terms from the ontology should be documented. Furthermore, Uschold and King point out that a large obstacle to efficient ontology sharing is the absence of documentation or inadequate documentation.

Ontology Development Methodologies – Grueninger and Fox

As you know, I am somewhat enamoured of ontologies and recently had cause to attend a meeting, where methods for ontology development and engineering were part of the discussion.

Now in software development, there are a number of methodologies, the community has converged on, such as the Capability Maturity Model (CMM), Waterfall Models and ISO standards, such as ISO 15504 and even ISO 9000.

Unfortunately this is not the case in ontological engineering where a number of different methodologies exist at the moment at the conceptual level (i.e. there is currently little to no tool support implementing these engineering approaches). So i thought I should write a series of blog posts, which will discuss some of the most common methodologies in ontological engineering in the hope that someone out there will find it useful (in the Long Tail I trust). Incidentally, you can tell I am on a train once more – I only get this sort of more fundamental high-level stuff done when travelling.

So first up is the methodology published bye Gueninger and Fox [1]. In pictorial form, the development process can be represented as follows:

Grueninger and Fox argue, that ontologies are constructed in response to an application or a particular problem, which is to be solved and hence a ontological engineering should begin with defining these problems in order to define an ontology’s scope. This should take the form of scenarios and questions that one might wish to ask of the ontology and which it should be capable of answering. In this sense, the questions define the compentency of the ontology and Grueninger and Fox have therefore coined the term “competency question”.

In a subsequent step, the terms of the ontology are defined. This includes objects, attributes and relationships. In essence, this defines the language the ontology will use and also provides semantic constraints. This is followed by the formal specification of the ontology (in formal language).

In the last step, the competency of the ontology is formally tested. Let’s look at all of these stages in greater detail.

Motivating Scenarios

As already stated, the basic assumption here is, that ontologies are constructed in response to a particular problem or requirement that might arise within the context of an application scenario. It can be argued, that quite often, the scenarios can be expressed in the form of story problems which can be written down and which not only often provide implicit solutions, but also allow a first and informal glimpse of the semantics of objects and relations, which will subsequently be included in the formal ontology.

Informal Competency Questions

Quite often, the consequence of the preparation of story problems and motivating scenarios will be the emergence of questions, that the ontology should be able to answer. These questions can be considered to represent queries over the ontology, which, at this stage have not been formalised. As such, they define the scope and therefore compentency of the ontology. Furthermore, the competency questions not only define the scope of the ontology, they also provide a good deal of the justification for building it in the first place and furthermore provide a first informal evaluation: can all the questions that ought to be asked of an ontology be answered within the ontology’s envisaged competency or is an extension required? As Grueninger and Fox point out, “[…]competency questions do not generate ontological commitments; rather, they are used to evaluate the ontological commitments that have been made.”

Formalisation of the Ontology

Once the scope and semantics of the ontology have been defined in this way, it is time for the formalisation of the ontology, i.e. the expression and definition of the ontology’s terminology in the form of objects, attributes and relationships in formal language. A large number of ontology languages can in principle be chosen from, though most modern ontoogies are constructed in either OBO or OWL.

Formal Competency Questions

The nest step is the formalisation of the competency questions in the form of entailments or consistency questions with respect to the axioms contained in the ontology. The formalisation of competency questions is an important part of the evaluation of the constructed ontology. It is also their main distinguishing feature: differentiation of ontologies occurs through the competency questions that different ontologies can answer. It is important to realise at this stage, that formal competency questions do no axiomatize the ontology, but will be used to evaluate its completeness.

Specification of Axioms

The definition of axioms is probably the most difficult part of the ontolgy engineering process. However, it the axioms in the ontology, which both specify the definition of terms as well as constrain their interpretation. In other words, it is the axioms, that provide the semantics of the ontology terms.
As Grueninger and Fox point out, axioms are central for ontology construction: “[…]without the axioms, we cannot express the question or its solution, and with the axioms we can express the question and its solutions.” Solutions to the competency questions must be entailed by the axioms in the ontology. If inconsistencies are found or formal competency questions cannot be answered, the ontology must be changed or extended until everything is satisfactory.

Completeness

In a last step, the ontology engineer must provide the conditions under which the solutions to formally stated competency questions are considered to be complete. This provides the basis for the formulation of completeness theorems.

Thus far Grueninger and Fox and half a knackered laptop battery later. Stay tuned for the next exciting installment of……..Uschold and King….:-).

[1] M. Grueninger, M. Fox, “Methodology for the Design and Evaluation of Ontologies”, IJCAI’95, Workshop on Basic Ontological Issues in Knowledge Sharing, April 13, 1995, url = citeseer.ist.psu.edu/grninger95methodology.html

Polymers with Mechanophores

Stimulus-responsive polymers are all the rage at the moment. pH, temperature, electrical stimuli etc.. have all been used to get a polymer to respond to its environment. And now, well, now there are mechanical stimuli too.

In a recent JACS communication (DOI)Moore et al. describe the synthesis of mechanophore-linked addition polymers.[1] Mechanophores are stress-sensitive units and through application of a stress, a chemical reaction is accelerated. Typical mechanophores, for example are Benzocyclobutene (InChI=1/C8H8/c1-2-4-8-6-5-7(8)3-1/h1-4H,5-6H) (BCB) or spiropyrans, which under go 4-pi or 6-pi stress-induced ring opening reactions.

Moore et al prepared addition polymers of the type PMA-BCB-PMA, starting from 1,2-bis(alpha-bromopropionyloxy)-1,2-dihydrobenzocyclobutene, using single electron transfer living radical polymerisation (SET-LRP):

Low (18 kDa), medium (91 kDa) and high (287 kDa) molecular weight polymers were synthesized, with PDIs of around 1.3. The ring-opening of the of the polymer was subsequently investigated by trapping the intermediate with N-(1-pyrene)maleimide as shown in step b of the above scheme (the pyrene moiety acts as a UV reporter).

To investigate the ring-opening reaction, the authors dissolved the polymer in acetonitrile together with the maleimide trap and a radical trap (2,6-di-tert-butyl-4-methylphenol, BHT) and subjected the solutions to pulsed sonication for 45 min under inert conditions. Analysis of the solution by UV spectroscopy showed, that no reaction had taken place in the low molecular weight polymer, but that both the medium and high-molecular weight polymers gave rise to significant UV signals, indicating that the maleimide is indeed incorporated via an electrocyclic ring opening process. Similar results were obtained for a high-molecular weight (170 kDa) PMA-spyropyran polymer:

When a solution of the colourless polymer was subjected to pulsed sonication, the solution changed colour to a pink hue with a new UV band at 550 nm. When the solution was exposed to ambient light at room temperature, the colour disappeared, which was found to be consistent with a known photolytic reversion to the closed form.

So there you have it, folks…..polymers responding to mechanical stimuli. Exciting work.

[1] Potisek, S. L. et al., J. Am. Chem. Soc., 129(45), 13808-13809 (2007)

Update (more InChIs by request)

1,2-bis(alpha-bromopropionyloxy)-1,2-dihydrobenzocyclobutene: InChI=1/C14H14Br2O4/c1/h3-8,11-12H,1-2H3/t7?,8?,11-,12+

Me₆Tren: InChI=1/C12H30N4/c1-13-13/h7-12H2,1-6H3
N-(1-pyrene)maleimide: InChI=1/C20H11NO2/c22-17-10-11-18(23)21(17)16-9-7-14-5-4-12-2-1-3-13-6-8-15(16)20(14)19(12)13/h1-11H

Screencasts for scientists.

So you have just installed a new piece of software you need to go about your work. You fire the thing up for the first time. Getting that sinking feeling already? Is the software in keeping with the fine traditions of academic software which only a minimalist or non-existent user interface? Software that has been written with the “user-friendliness is for wimps” mantra and is as usable as a blended iPod? Gromacs keeping you up at night? And the only way to cope with it all, is to work through large heaps of documentation (if existent?).

Well, there may be a solution on the horizon: Bioscreencast. Screencast is a a free video-tutorial library specifically for science-related software tools. The number of tutorials is small so far (the site is in beta), but it contains all sorts of useful things from how to use BLAST and sequence alignment tools to PyMol tips and tricks and ideas for the “paperless PhD”:

Much to my delight I have also discovered a screencast on how to model physical objects in OWL:

and there are further tutorials on how to use Reference Managers, Connotea etc. Users can also request tutorials (one request is for an R tutorial (much needed I think), other users can vote for the request and someone can pick up the baton and produce a tute.

Furthermore, the site provides all the necesary functionality to capture a screencast: no software downloads etc….it’s a little bit like Seesmic. It seems like a great way to show off (new) software and tools, to demonstrate how to use them and to make a contribution to the community.

I will certainly have a think what we can contribute…and I would urge you to check out the site.

More of micelles and porins….and symmetry.

A while ago I blogged about the incorporation of porins, channel-forming proteins, into synthetic polymer membranes and in particular into micelles and the binding of phages to the incorporated porins.

I did not mention that this is usually done through simple diffusion processes: one can either preform the membrane over a hole on a substrate and add a drop of a solution of porins to this or one mixes the block-copolymer solution and the porin prior to forming micelles. While this is a well-established way of doing things and has been demonstrated to be effective in a number of experiments, it is not very good when trying to imitate biological systems.

Why? Well, proteins have both an amino-terminus and a carboxy terminus. In nature, porins are “vectorial” molecules, they have very distinct intracellular and extracellular parts. In “real” biological systems, the amino-terminus of a porin is located in the cytoplasm, i.e. inside the cell. Phages, however, typically bind to the carboxy-terminus, i.e. the part of the protein outside the cell. Incorporation of porins in the synthetic membranes by diffusion is essentially a statistical process and hence one would expect only half of the porins to be incorporated into a membrane in the physiologically correct orientation. So what to do? Well, the structure of natural membranes may hold a clue.

Meier and colleagues noted, that a lot of physicochemical and immunological properties of cell membranes are dependent on the lipid composition of both membrane leaflets and that this composition is usually asymmetric.(DOI).[1] They furthermore noted, that all of the polymers used in membrane/porin experiments are symmetric with respect to their mid-planes if one disregards curvature effects for a moment. For this reason, they prepard both a symmetric poly(2-methyl-2-oxazoline)-block-poly(dimethyl siloxane)-block-poly(2-methyl-2-oxazoline) (PMOXA-PDMS-PMOXA) copolymer, as well as asymmetric poly(ethylene oxide)-block-poly(dimethyl siloxane)-block-poly(2-methyl-2-oxazoline) (PEO-PDMS-PMOXA) polymers. The latter was present in two forms containing both a large (PEO₂₅-PDMS₄₀-PMOXA₁₁₀) and a small (PEO₆₇-PDMS₄₀-PMOXA₄₅) poly(2-methyl-2-oxazoline) block.

All of the polymers form vesicles when dissolved in water, with the hydrophobic block being covered by the hydrophilic blocks on both the outside and the inside of the vesicle wall with the more voluminous hydrobilic block on the outside of the micelle. For this reason, the PEO₂₅-PDMS₄₀-PMOXA₁₁₀ polymer is expected to give rise to an ABC motive, and PEO₆₇-PDMS₄₀-PMOXA₄₅ to a CBA orientation.

Figure 1: The asymmetric unit of aquaporin0.

Aquaporin0, labelled with histidine tags on the amino-terminus was subsequently incorporated into the vesicles formed by all three polymers. An antibody assay, using an antibody specific to histidine, showed that in the case of the symmetric PMOXA-PDMS-PMOXA polymer, the aquaporin did indeed incorporate statistically into the vesicle membrane, with an equal amount of histidine tags located on the outside and inside walls of the vesicle wall. The ABC motif, by contrast, led to a physiological incorporation of the porins, with approximately 80 % of the histidine tags inside the vesicle (corresponding to the cytoplasm in a cell). For the CBA motif, the situation was reversed: 70 % of the histidine residues are now located on the outside of vesicle walls.

In this way the authors have clearly demonstrated, that breaking the symmetry of a synthetic membrane system results in the directed insertion of membrane proteins.

[1] Stoenescu, R.; Graff, A.; Meier, W. Macromol. Biosci. 2004, 4, 930-935