Conceptual Modeling Applied to Genomics: Challenges Faced in Data Loading

Abstract

Todays genomic domain evolves around insecurity: too many imprecise concepts, too much information to be properly managed. Considering that conceptualization is the most exclusive human characteristic, it makes full sense to try to conceptualize the principles that guide the essence of why humans are as we are. This question can of course be generalized to any species, but we are especially interested in this work in showing how conceptual modeling is strictly required to understand the ''execution model'' that human beings ''implement''. The main issue is to defend the idea that only by having an in-depth knowledge of the Conceptual Model that is associated to the Human Genome, can this Human Genome properly be understood. This kind of Model-Driven perspective of the Human Genome opens challenging possibilities, by looking at the individuals as implementation of that Conceptual Model, where different values associated to different modeling primitives will explain the diversity among individuals and the potential, unexpected variations together with their unwanted effects in terms of illnesses.

This work focuses on the challenges faced in loading data from conventional resources into Information Systems created according to the above mentioned conceptual modeling approach. The work reports on various loading efforts, problems encountered and the solutions to these problems. Also, a strong argument is made about why conventional methods to solve the so called `data chaos¿ problems associated to the genomics domain so often fail to meet the demands.