A Methodology for the Development of RESTful Semantic Web Services for Gene Expression Analysis.

Gabriela D A Guardia,Kelen C R Malmegrim,Luís Ferreira Pires,Cléver R G De Farias,Ricardo Z N Vêncio,Avi Ma’Ayan

doi:10.1371/journal.pone.0134011

Abstract

Gene expression studies are generally performed through multi-step analysis processes, which require the integrated use of a number of analysis tools. In order to facilitate tool/data integration, an increasing number of analysis tools have been developed as or adapted to semantic web services. In recent years, some approaches have been defined for the development and semantic annotation of web services created from legacy software tools, but these approaches still present many limitations. In addition, to the best of our knowledge, no suitable approach has been defined for the functional genomics domain. Therefore, this paper aims at defining an integrated methodology for the implementation of RESTful semantic web services created from gene expression analysis tools and the semantic annotation of such services. We have applied our methodology to the development of a number of services to support the analysis of different types of gene expression data, including microarray and RNASeq. All developed services are publicly available in the Gene Expression Analysis Services (GEAS) Repository at http://dcm.ffclrp.usp.br/lssb/geas. Additionally, we have used a number of the developed services to create different integrated analysis scenarios to reproduce parts of two gene expression studies documented in the literature. The first study involves the analysis of one-color microarray data obtained from multiple sclerosis patients and healthy donors. The second study comprises the analysis of RNA-Seq data obtained from melanoma cells to investigate the role of the remodeller BRG1 in the proliferation and morphology of these cells. Our methodology provides concrete guidelines and technical details in order to facilitate the systematic development of semantic web services. Moreover, it encourages the development and reuse of these services for the creation of semantically integrated solutions for gene expression analysis.

Highlights

Gene expression analysis has become a powerful tool for acquiring new knowledge about the mechanisms involved in a number of biological processes and diseases
Each RESTful web service should expose a set of resources, i.e., any type of information exclusively identified by a Uniform Resource Identifier (URI) that can be referenced through the web
This section presents our methodology for the systematic development of RESTful semantic web services from existing software tools. This methodology consists of four general activities: (i) create a RESTful web service from an existing software tool; (ii) automatically generate a WSDL service description from the service implementation; (iii) define domain and service ontologies; and (iv) use these ontologies to semantically annotate the WSDL service description according to the Semantic Annotations for WSDL (SAWSDL) approach

Summary

Introduction

Gene expression analysis has become a powerful tool for acquiring new knowledge about the mechanisms involved in a number of biological processes and diseases. Gene expression analysis usually demands the integrated use of a number of analysis tools. The integration of a series of tools to carry out an analysis process requires users to manually transfer data between different tools and/or to convert data formats due to data structural differences. This process can be time consuming and error prone due to the increasing number of analysis tools and data formats available in this domain [1, 2]. The availability of integrated environments is highly desirable in this domain

Objectives

Methods

Results

Discussion

Conclusion