Network-Based Methods for Computational Diagnostics by Means of R
Networks representing biomedical data have become a powerful approach in different research disciplines dealing with complex diseases. Also, R and Bioconductor have emerged as a standard research environment to investigate and analyze high-throughput data
- PDF / 251,936 Bytes
- 13 Pages / 504.567 x 720 pts Page_size
- 6 Downloads / 199 Views
11
Laurin A.J. Mueller, Matthias Dehmer, and Frank Emmert-Streib
Abstract
Networks representing biomedical data have become a powerful approach in different research disciplines dealing with complex diseases. Also, R and Bioconductor have emerged as a standard research environment to investigate and analyze high-throughput data. Therefore, we present and discuss existing packages, available in R or Bioconductor, that provide methods for computational diagnostics by means of networks. In particular, we summarize packages to reconstruct and analyze networks from high-throughput data. Moreover, we discuss packages that provide comprehensive methods to visualize large-scale gene networks in order to support the field of computational diagnostics of complex diseases. The aim of this chapter is to support an interdisciplinary research community dealing with computational diagnostics to investigate novel hypothesis in a medical and clinical context to gain a better understanding of complex diseases.
11.1
Introduction
During the last 15 years, networks have gained a considerable interest throughout a variety of different scientific disciplines (Bornholdt and Schuster 2003). Initiated by the seminal work of Watts and Strogatz (1998) and Baraba´si and Albert (1999), who showed that the network structure found in real networks is considerably different from random networks, nowadays, considering a problem
M. Dehmer (*) UMIT, Institute for Bioinformatics and Translational Research, Eduard Wallnoefer Zentrum 1, 6060 Hall in Tyrol, Austria e-mail: [email protected]
from a network perspective is well established. For this reason it is of no surprise that also in biology and the biomedical sciences networks and networkbased approaches are omnipresent. One reason for this is that it has been recognized that a systems approach (von Bertalanffy 1950) can be mathematically realized by a network approach, because networks represent naturally the interactions among genes or gene products (Emmert-Streib and Dehmer 2011). Due to the fact that it is generally acknowledged that interactions among genes are responsible for the emergence of a phenotype of an organism, network-based approaches enable the investigation of the biological function of biological cells (Vidal 2009; Waddington 1957). Within the last couple of years high-throughput technologies, e.g., proteomics, DNA microarray or
Z. Trajanoski (ed.), Computational Medicine, DOI 10.1007/978-3-7091-0947-2_11, # Springer-Verlag Wien 2012
185
186
L.A.J. Mueller et al.
next-generation sequencing, originating from basic biological research, are more and more used in a medical and even clinical context. Data generated from complex diseases are holding great promise to provide a valuable source for novel diagnostic methods that bear the potential to revolutionize medical research because they provide genomewide information about thousands of genes and their interactions. Hence, the generation of highthroughput data of various types and their integration could trigger dat
Data Loading...
