Content uploaded by Sacha A F T van Hijum
Author content
All content in this area was uploaded by Sacha A F T van Hijum on Aug 19, 2014
Content may be subject to copyright.
Available via license: CC BY-NC 2.0
Content may be subject to copyright.
Vol. 23 no. 9 2007, pages 1161–1163
BIOINFORMATICS APPLICATIONS NOTE doi:10.1093/bioinformatics/btl658
Gene expression
FIVA: Functional Information Viewer and Analyzer extracting
biological knowledge from transcriptome data of prokaryotes
Evert-Jan Blom
1
, Dinne W. J. Bosman
2
, Sacha A. F. T. van Hijum
1
, Rainer Breitling
3
,
Lars Tijsma
2
, Remko Silvis
1
, Jos B. T. M. Roerdink
2
and Oscar P. Kuipers
1,
1
Molecular Genetics, Groningen Biomolecular Sciences,
2
Institute for Mathematics and Computing Science and
3
Groningen Bioinformatics Centre, University of Groningen, PO Box 800, 9700 AV, Groningen, The Netherlands
Received on October 19, 2006; revised on November 24, 2006; accepted on December 19, 2006
Advance Access publication January 19, 2007
Associate Editor: Dmitrij Frishman
ABSTRACT
Summary: FIVA (Function Information Viewer and Analyzer) aids
researchers in the prokaryotic community to quickly identify relevant
biological processes following transcriptome analysis. Our software
assists in functional profiling of large sets of genes and generates
a comprehensive overview of affected biological processes.
Availability: http://bioinformatics.biol.rug.nl/standalone/fiva/
Contact: o.p.kuipers@rug.nl
Supplementary information: http://bioinformatics.biol.rug.nl/stand
alone/fiva/suppMaterials.php
1 INTRODUCTION
Genome-wide expression profiles describing various cellular
states are obtained by use of DNA microarrays. Following
statistical analysis of the raw gene expression values, data-
driven methods such as unsupervised clustering allow grouping
of genes based on their (temporal) expression patterns. Genes
involved in similar cellular processes are expected to have a
high probability of exhibiting similar expression patterns.
Analysis and interpretation of these clusters is time-consuming
and error-prone. Various applications have been developed to
functionally profile differentially expressed genes from DNA-
microarray experiments.
Several of these, as reviewed by Khatri et al. (2005), overlap
with our application in terms of functionality and data
sources employed. Many of these focus on higher organisms
and therefore lack support for prokaryote gene identifiers.
A number of applications support rarely used (Uniprot,
GI accession) identifiers (Hosack et al., 2003) or only identifiers
for a limited set of organisms (Scheer et al., 2006). Moreover,
with few exceptions, these software products use gene ontology
as their exclusive data source. In addition, the laborious task
of preprocessing the list containing differentially expressed
genes must be performed by a researcher. A stand-alone
application that focuses on prokaryotes is therefore essential
for the fast-growing community of microbiologists making use
of a plethora of (confidential) microbial genome sequences.
We have developed FIVA (Functional Information Viewer
and Analyzer). It uses several sources of biological annotations
to create an extensive functional profile based on gene
expression data. Furthermore, FIVA is capable of processing
groups of genes assembled by other criteria (e.g. functional
grouping of genes which are not available in current annotation
modules). The significance of each biological process is
calculated to distinguish between significant and spurious
occurrences.
2 PROGRAM OVERVIEW
2.1 Input
The input data for FIVA consists of transcriptome data and
genome annotation files (e.g. EMBL or Genbank), supplemen-
ted with annotation information. FIVA supports a broad
variety of prokaryotic gene identifiers from the expression
datasets, including locus tags and standard gene names (further
details available in Supplementary Materials). Each annotation
module uses functional information from one of the following
sources to classify the groups of genes and determine any
significantly over-represented categories. (i) Gene ontology
(ii) Metabolic pathways (iii) COG classes (iv) Regulatory
interactions (v) UniProt keywords (vi) InterPro (vii) User-
defined functional categories.
2.2 Processing
The analysis in FIVA first involves the partitioning of the gene
expression data into up- and down-regulated fractions. Testing
different settings to partition the data is not a trivial task.
FIVA offers the ability to automatically detect the optimal
settings for each individual experiment based on the number of
over-represented functional categories. In addition to this
partitioning method which is based on thresholds applied to
a single experiment, the iGA algorithm (Breitling et al., 2004) is
implemented. This algorithm optimizes the parameters for each
functional category, which greatly improves the sensitivity of
the analysis and increases the number of affected biological
processes that can be reliably detected. Furthermore, the
analysis of gene expression data can also be applied on
user-defined gene lists.
*To whom correspondence should be addressed.
ß2007 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/
by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
by guest on June 4, 2013http://bioinformatics.oxfordjournals.org/Downloaded from
A Fisher exact test is used to calculate P-values for each
cluster. This P-value describes the probability of observing a
specific enrichment of genes from a functional category in
a cluster by chance. The number of false positives, due to the
large number of statistical tests performed, are controlled
by four multiple testing corrections (Benjamini/HochBerg,
Bonferroni Step-down, Bonferroni and Benjamini Yekutieli).
These are implemented to adjust the raw P-values (see
Supplementary Website).
2.3 Output
For each of the classification modules, a graphical representa-
tion of the over-represented categories is generated. A preview
is created from these results, from which a selection of the
results can be made (Fig. 1). In order to conveniently compare
biological phenomena occurring in different experiments,
multiple experiments can be loaded simultaneously and are
displayed as columns. Clickable links are present for each
category in the individual graphical map, providing detailed
information for each cluster that contains an enrichment of
genes from this category. Furthermore, FIVA uses the KEGG
API (http://www.genome.jp/kegg/soap/) to communicate with
the KEGG database to color pathways based on the gene
distribution in the clusters.
2.4 Implementation and availability
FIVA was programmed as a stand-alone application in Java
using the Eclipse (http://www.eclipse.org/) framework and
runs on all Java-supporting operating systems (Mac OS,
MS Windows, UNIX and Linux). The graphical output can
also be viewed by all web browsers that are able to process
scalable vector graphics (SVG) or, to ensure portability
Fig. 1. Graphical output of a single annotation module. Genes from two DNA-microarray datasets (gluc: growth on glucitol compared to growth on
glucose, man: growth on mannitol compared to glucose) were partitioned into up- and down-regulated clusters. The size of each cluster is displayed
in blue underneath the cluster name. Numbers in each rectangle represent absolute values of occurrences. The significance of occurrences is visualized
in a colour gradient which is displayed at the bottom of the plot. The description of each category is placed at the right. S: annotations that are
significant after multiple testing correction. Multiple testing correction results are visualized using five different symbols to distinguish between the
individual corrections. The number of symbols placed in each rectangle corresponds to the number of multiple testing corrections after which the
annotation is found significant.
E.-J.Blom et al.
1162
by guest on June 4, 2013http://bioinformatics.oxfordjournals.org/Downloaded from
of the results, portable network graphics. More information
on the functionality of FIVA, as well as the results of
several test cases, can be found under the Supplementary
Materials.
3 CONCLUSION
A full information analysis was performed to assess the
overlap between the annotation modules (see Supplementary
Website). The gene ontology module is the most informative
annotation type for our test organism Bacillus subtilis and
covers a large portion of the information present in the other
types. However, the utilization of multiple modules yields
relevant areas which are not shared by any of the other
modules. For our test cases, several relevant categories were
identified by the metabolic pathways modules but were missed
by the GO module (see Supplementary Website for more
information on this analysis). We conclude that combining
multiple annotation sources into one tool is advantageous
compared to using only one or a few sources. The combination
of various complementary annotation sources, together
with the dynamic visualization and elaborate statistical
analysis, allows a richer and more objective exploration
of prokaryote expression data than any other available tool
provides.
ACKNOWLEDGEMENTS
This study was fully supported by a grant from The
Netherlands Organization for Scientific Research and industrial
partners in the NWO-BMI project number 050.50.206 on
Computational Genomics of Prokaryotes and by Center IOP
Genomics. Work performed by SvH was supported by grant
QLK3-CT-2001-01473 under the EU programme ‘Quality of
life and management of living resources: The cell factory’.
We thank J.W.Veening for useful suggestions on experimental
procedures and G. te Meerman for expert advice on the
statistical analysis. Funding to pay the Open Access charges
was provided by the Molecular Genetics department of the
University of Groningen.
Conflict of Interest: none declared.
REFERENCES
Breitling,R. et al. (2004) Iterative group analysis (iGA): a simple tool to enhance
sensitivity and facilitate interpretation of microarray experiments. BMC
Bioinformatics,5, 34.
Hosack,D.A. et al. (2003) Identifying biological themes within lists of genes with
ease. Genome Biol.,4, R70.
Khatri,P. et al. (2005) Ontological analysis of gene expression data: current tools,
limitations, and open problems. Bioinformatics,21, 3587–3595.
Scheer,M. et al. (2006) JProGO: a novel tool for the functional interpretation
of prokaryotic microarray data using Gene Ontology information.
Nucleic Acids Res.,34, W510–515.
FIVA: Functional Information Viewer and Analyzer
1163
by guest on June 4, 2013http://bioinformatics.oxfordjournals.org/Downloaded from
