4th European Conference of Philosophy of Science, 28-31 August 2013, Helsinki
Symposium: Models in Bioinformatics
Room 7, Friday 15:30-17:30
Chair: Federica Russo
Speakers:
Emanuele Ratti, European School of Molecular Medicine, Milan
Federico Boem, European School of Molecular Medicine, Milan
Emilio M. Sanfillippo, (former) Institue for Formal Ontology and Medical Information Science, Saarbruecken; ISTC-CNR Laboratory for Applied Ontology, Trento
Abstract of the symposium
Models are usually considered as artefacts for the representation of some aspects of the real world that are relevant for the modeller’s purposes according with available knowledge. Modeling is an integral part of different scientific research initiatives, mainly because they facilitate the study of specific phenomena by simplifying or reducing them to some particular and relevant aspects. In Philosophy of Science little attention has been paid so far to the critical analysis of modeling in disciplines like Systems- and Molecular Biology, which are playing the role of leading researches initiatives. In this symposium we are interested in analysing the exploitation of computational models in these fields, i.e. those artefacts representing biological phenomena using mathematical and computational methods.
The growing importance of computer based modelling in Systems- and Molecular Biology is due to the fact that they need to process big portions of data in order to perform experiments. Emerging disciplines as Computational Biology and Bioinformatics are thus assuming a central role, by applying formal methods to the study of biological processes and entities. Their importance is due to the fact that by using engineering and computational approaches, the chances of performing more elaborated experiments and of getting more elaborated results are drastically improved.
We think that the philosophical debate on scientific models can be enlarged by the analysis of Bioinformatics strategies regarding models development, integration, and evaluation. The aim of the symposium is thus the analysis of certain issues related to the debate on scientific models in Philosophy of Science through case studies mainly taken from Bioinformatics.
We first analyse what Bioinformatics models are, comparing them with bio-ontologies. The contemporary application of formal ontological theories to Computer Science has indeed launched a new scientific enterprise, called Applied Ontology, or Ontology Engineering. By using ontologies, models can be consistently developed and eventually harmonized to a general view of reality. However, the notions of ‘model’ and ‘ontology’ have to be clearly distinguished and for doing this we take into account mainly studies from Applied Ontology for biomedical applications.
Secondly, we focus on the notion of idealization and integration in modelling by case studies from Networks Biology. The idea is that the process of integration of big data within Networks Biology databases is a special case of idealization as conceived in the debate of scientific models within Philosophy of Science. Integration is indeed a useful tool to make hypotheses that could not be made otherwise due to the big amount of data scientists have to deal with. In philosophical terms, integration allows surrogative reasoning over big data.
Last but not least, we examine the realist approach to modelling, while proposing at the same time its critique on the basis of referentialist semantic theory of models. We mean that computational ontologies run the risk of facing the same semantic problems envisaged by logical neo-empiricists concerning the supposed isomorphism between material world and reference language. In order to avoid such a situation we clarify some assumptions about bio-ontologies and their use in ITC-based biomedical research.
Keywords: scientific models, formal semantics, bio-ontology, Bioinformatics, Applied Ontology, Network Biology
Program and Book of abstracts available: at: http://www.helsinki.fi/epsa13/Materials/EPSA13%20Abstract%20Book.pdf
