Minimum Ontology for 2DE Gel Electrophoresis

This document is developed to serve as a human guide on how to use The Charleston Core Minimum Ontology for 2DE Gel Electorphoresis, herein refers to MO2DE. MO2DE is an ontology designed for annotating information necessary to describe the production of a 2DE experiment. In particular, the ontology is designed as the RDF counterpart of Minimum Information about 2DE Gel (MI2DG).

Ontology design differs from an XML schema design in that the former is a process of modelling knowledgewhereas the later is of data. Any modelling process must employ, and therefore comply to the constraints, of a programming language. To model data in XML, one must model its data into a tree. However, not all data are necessarily expressed as a tree; nor can all of them. Therefore, a data structure in a XML schema sometimes is designed not for the semantic relationshp of data but for practical, sometimes aesthetic, purposes. Modeling knowledge, on the other hand, is different. First, the language to be used OWL/RDF employes a directed labeled graph (DLG). Second, ontology design should consider the potential (or expected) use of inference engine. So, the modeling should adhere to its semantic relationship as much as possible.

However, an all-RDF solution will make an ontology to resemble very little to its XML counterpart. The design of this ontology therefore takes a middle road. At one hand, we try to ground it to existing ontologies, for instance, BOSS; On the other hand, we tried to mimic data relationship of its XML counter part as much as we can.

The RDF definition of MO2DG is located at http://www.charlestoncore.org/ontology/mo2dg

A 2DE Study

In general, MO2DG considers a 2DE Gel work be a boss:Study. Every TwoDEStudy shall use some BioSample as boss:material and and GelProtocol as boss:method (see Figure 1).

Overview of MO2DG

Figure 1: Overview of MO2DG in DLG² .

Sample

A BioSample is the subject of study. Each BioSample must label its species, tissue and cell (see Figure 2).

Biological Sample

Figure 2: Design of BioSample in DLG²..

Note

Here species, tissues and cell is specified as an xsd:string. A more RDF-like way shall use a URI. But we do not think we are in the position to assign this URI so using xsd:string is a temporary solution. But meanwhile, we recomment using controlled vocabularies (for example, MeSH) as much as possible.

Gel Protocol

A typical 2DE gel experiment follows the following steps:

Prepare sample
Load sample
Run first dimension, typically isoelectric focusing
Run second dimension, typically SDS-PAGE
Stain the gel
Digitize the gel

The above six steps are modelled into the ontology structure shown in Figure 3. Among them, only three of them are modeled as a class of boss:Protocol. Two of them - staining and densitomety - are simply modeled as a string element. As we mentioned earlier, such a treatment is to accommodate the design of its XML counterpart. We will discuss a more RDF-like treatment in a future document.

Figure 3: GelProtocol

Figure 3: Design of GelProtocol in DLG² .

Sample Preparation

Preparing samples for a 2DE gel usually involves two main processing steps. They are modeled as shown in Figure 4.

Prefractionation: to reduce the complexity of samples and enrich proteins of interest. Such step is modeled as a owl:DatatypeProperty of xsd:string.
Solubilization: in order to obtain the consistent molecular weight and pI of the proteins of interest, all noncovalent bounds should be disrupted. Such steps are usually achieved by treating samples in some lysis buffer that contains chaotropes, detergents, reducing agent, ampholyte, protease inhibitors and alklynation agents.

Figure 4: Sample Preparation

Figure 4: Design of SamplePrep in DLG² .

Isoelectric Focusing Protocol

The first dimension of a 2DE gel usually involves separating protein mixtures according to their isoelectric point. The separation is done by applying electric field to a medium of certain pH gradient. This process is modeled as shown in figure 5.

Figure 5: Isoelectric Focusing Protocol

Figure 5: Design of IEF_Protocol in DLG² .

SDS-Page Protocol

The separation of protein on the second dimension of a 2DE is by the size of proteins. Such step can be achieved by applying electric field on SDS polyacrylamide gel. Such step is modeled in Figure 6.

Figure 6: SDS-PAGE Protocol

Figure 6: Design of SDS-PAGE Protocol in DLG² .

Reagent

A biological experiment need chemical reagents. The Reagent class is created for the purpose. Once again, this design is trying to accommodate its XML counter part. A more reasonable approach will at least separate them into a separate namespace. The desing of Reagent and some classes used in this ontology is shown in Figure 7.

Figure 7: Reagent

Figure 7: Design of Reagent in DLG² .

Additional Note

Many concepts existed in MI2DG are not designed in MO2DG becaue first, it is not necessary. For instance, "mi2dg/protocol_id" is not needed because RDF by default requires a URI to make sense. Second, many concepts that have been developed elsewhere. For example, most of the "description", "note", "date", "submitter" etc. can use "description", "date", "creator" developed by Dublin Core Metadata Initiative.

Classes

TwoDEStudy

The general concept used to describe a study using 2DE gel electrophoresis as its methdology.

Constraints: A TwoDEStudy must use at least one BioSample as its boss:material and one GelProtocol as its bossmethod.

Conterparts in MI2DG: "/mi2dg"

BioSample

The biological materials that will be processed and run on a gel during a TwoDEStudy.

Constraints: A BioSample has three mandatory properties: species, tissue and cell, all of which are of type xsd:string.

Conterparts in MI2DG: "/mi2dg/sample_type"

Reagent

The concept is used to represent chemicals used in an experiment protocols.

Constraints: A Reagent must be provided with the following three properties: chemName, concentration and unit, each of which is an owl:DatatypeProperty of xsd:string.

Counterpart in MI2DG: none.

Chaotrope

Chaotrope refers to the chemical reagents, like high concentration of urea, that are used to destablize proteins.