BRASIL
BRASIL
PeerJ, 5, e3509.
Abstract:
There are many steps in analyzing transcriptome data, from the
acquisition of raw data to the selection of a subset of representative
genes that explain a scientific hypothesis. The data produced can be
represented as networks of interactions among genes and these may
additionally be integrated with other biological databases, such as
Protein-Protein Interactions, transcription factors and gene annotation.
However, the results of these analyses remain fragmented, imposing
difficulties, either for posterior inspection of results, or for
meta-analysis by the incorporation of new related data. Integrating
databases and tools into scientific workflows, orchestrating their
execution, and managing the resulting data and its respective metadata
are challenging tasks. Additionally, a great amount of effort is equally
required to run in-silico experiments to structure and compose the
information as needed for analysis. Different programs may need to be
applied and different files are produced during the experiment cycle. In
this context, the availability of a platform supporting experiment
execution is paramount. We present GeNNet, an integrated transcriptome
analysis platform that unifies scientific workflows with graph databases
for selecting relevant genes according to the evaluated biological
systems. It includes GeNNet-Wf, a scientific workflow that pre-loads
biological data, pre-processes raw microarray data and conducts a series
of analyses including normalization, differential expression inference,
clusterization and gene set enrichment analysis. A user-friendly web
interface, GeNNet-Web, allows for setting parameters, executing, and
visualizing the results of GeNNet-Wf executions. To demonstrate the
features of GeNNet, we performed case studies with data retrieved from
GEO, particularly using a single-factor experiment in different analysis
scenarios. As a result, we obtained differentially expressed genes for
which biological functions were analyzed. The results are integrated
into GeNNet-DB, a database about genes, clusters, experiments and their
properties and relationships. The resulting graph database is explored
with queries that demonstrate the expressiveness of this data model for
reasoning about gene interaction networks. GeNNet is the first platform
to integrate the analytical process of transcriptome data with graph
databases. It provides a comprehensive set of tools that would otherwise
be challenging for non-expert users to install and use. Developers can
add new functionality to components of GeNNet. The derived data allows
for testing previous hypotheses about an experiment and exploring new
ones through the interactive graph database environment. It enables the
analysis of different data on humans, rhesus, mice and rat coming from
Affymetrix platforms. GeNNet is available as an open source platform at https://github.com/raquele/GeNNet and can be retrieved as a software container with the command docker pull quelopes/gennet.
