Registering records into models

A quick primer on what models are and how to use them to capture patient and sample data in your biobank

Using data models helps you keep records organized, logically separated, and manageable. In this article, we will establish what data models are, how to best use them, and an example of data models in biobank scenarios.

Data Models

Simply put, a data model is a categorization technique. In the BioBox systems, models are a collection of conditions related to the properties of the data record. Taken together, these conditions creates the qualifiers that a record must have to be considered part of that model.

Creating Data Model Conditions


Using any of your dictionary terms, you can set up a condition within a model. Depending on the type of dictionary term, there may be different operators you can use.

Results of applying conditions

To illustrate the effect of these conditions, consider a hypothetical data model named "High Grade Glioma" that uses the following conditions:

Property Condition Value
Diagnosis Includes Glioma
WHO Grade Greater than 3

When applied to the following records, the membership results as follows:

Record Name Diagnosis WHO Grade Included/Excluded
P-001 Glioma 2 Excluded
P-002 Glioma 4 Included
P-003 Medulloblastoma 3 Excluded
P-004 Oligodendroglioma 4 Excluded

 

Records can belong to zero, one, or more models

 

Biobank setup

The simplest version of a biobank only needs two models; Patient and Sample.

Both of these models relies only on a single dictionary term: Record Class. The outcome of these two models, given the following records are:

Record Name Record Class Model Membership
P-001 Patient Patient
P-002 Patient Patient
S-001 Sample Sample
S-002 Sample Sample
U-001 (empty) (empty)