Answers to the most complex biological questions are rarely determined solely from the experimental evidence. It requires subsequent analysis of many data sources that are often heterogeneous. Most biological data repositories focus on providing only one particular type of data, such as sequences, molecular interactions, protein structure, or gene expression. In many cases, it is required for researchers to visit several different databases to answer one scientific question. It is essential to develop strategies to integrate disparate biological data sources that are efficient and seamless to facilitate the discovery of novel associations and validate existing hypotheses. This thesis presents the design and development of different integration strategies of biological and clinical systems. The BioSPIDA system is a data warehousing solution that integrates many NCBI databases and other biological sources on protein sequences, protein domains, and biological pathways. It utilizes a universal parser facilitating integration without developing separate source code for each data site. This enables users to execute fine-grained queries that can filter genes by their protein interactions, gene expressions, functional annotation, and protein domain representation. Relational databases can powerfully return and generate quickly filtered results to research questions, but they are not the most suitable solution in all cases. Clinical patients and genes are typically annotated by concepts in hierarchical ontologies and performance of relational databases are weakened considerably when traversing and representing graph structures. This thesis illustrates when relational databases are most suitable as well as comparing the performance benchmarks of semantic web technologies and graph databases when comparing ontological concepts. Several approaches of analyzing integrated data will be discussed to demonstrate the advantages over dependencies on remote data centers. Intensive Care Patients are prioritized by their length of stay and their severity class is estimated by their diagnosis to help minimize wait time and preferentially treat patients by their condition. In a separate study, semantic clustering of patients is conducted by integrating a clinical database and a medical ontology to help identify multi-morbidity patterns. In the biological area, gene pathways, protein interaction networks, and functional annotation are integrated to help predict and prioritize candidate disease genes. This thesis will present the results that were able to be generated from each project through utilizing a local repository of genes, functional annotations, protein interactions, clinical patients, and medical ontologies.

Made available in DSpace on 2016-06-02T19:05:56Z (GMT). No. of bitstreams: 1 4370.pdf: 1089392 bytes, checksum: 82daa0e51d37184f8864bd92d9342dde (MD5) Previous issue date: 2012-02-29 / In bioinformatics, there is a huge volume of data related to biomolecules and to nucleotide and amino acid sequences that reside (in almost their totality) in several Biological Data Bases (BDBs). For a specific sequence, there are some informational classifications: genomic data, evolution-data, structural data, and others. Some BDBs store just one or some of these classifications. Those BDBs are hosted in different sites and servers, with several data base management systems with different data models. Besides, instances and schema might have semantic heterogeneity. In such scenario, the objective of this project is to propose a biological data integration model, that adopts new schema integration and instance integration techniques. The proposed integration model has a special mechanism of schema integration and another mechanism that performs the instance integration (with support of a dictionary) allowing conflict resolution in the attribute values; and a Clustering Algorithm is used in order to cluster similar entities. Besides, a domain specialist participates managing those clusters. The proposed model was validated through a study case focusing on schema and instance integration about nucleotide sequence data from organisms of Actinomyces gender, captured from four different data sources. The result is that about 97.91% of the attributes were correctly categorized in the schema integration, and the instance integration was able to identify that about 50% of the clusters created need support from a specialist, avoiding errors on the instance resolution. Besides, some contributions are presented, as the Attributes Categorization, the Clustering Algorithm, the distance functions proposed and the proposed model itself. / Na bioinformática, existe um imenso volume de dados sendo produzidos, os quais estão relacionados a sequências de nucleotídeos e aminoácidos que se encontram, em quase a sua totalidade, armazenados em Bancos de Dados Biológicos (BDBs). Para uma determinada sequência existem algumas classificações de informação: dados genômicos, dados evolutivos, dados estruturais, dentre outros. Existem BDBs que armazenam somente uma ou algumas dessas classificações. Tais BDBs estão hospedados em diferentes sites e servidores, com sistemas gerenciadores de banco de dados distintos e com uso de diferentes modelos de dados, além de terem instâncias e esquemas com heterogeneidade semântica. Dentro desse contexto, o objetivo deste projeto de mestrado é propor um Modelo de Integração de Dados Biológicos, com novas técnicas de integração de esquemas e integração de instâncias. O modelo de integração proposto possui um mecanismo especial de integração de esquemas, e outro mecanismo que realiza a integração de instâncias de dados (com um dicionário acoplado) permitindo resolução de conflitos nos valores dos atributos; e um Algoritmo de Clusterização é utilizado, com o objetivo de realizar o agrupamento de entidades similares. Além disso, o especialista de domínio participa do gerenciamento desses agrupamentos. Esse modelo foi validado por meio de um estudo de caso com ênfase na integração de esquemas e integração de instâncias com dados de sequências de nucleotídeos de genes de organismos do gênero Actinomyces, provenientes de quatro diferentes fontes de dados. Como resultado, obteve-se que aproximadamente 97,91% dos atributos foram categorizados corretamente na integração de esquemas e a integração de instâncias conseguiu identificar que aproximadamente 50% dos clusters gerados precisam de tratamento do especialista, evitando erros de resolução de entidades. Além disso, algumas contribuições são apresentadas, como por exemplo a Categorização de Atributos, o Algoritmo de Clusterização, as funções de distância propostas e o modelo MIDB em si.