ral
ssBioMed CentBMC Bioinformatics
Open AcceProceedings
DoOPSearch: a web-based tool for finding and analysing common
conserved motifs in the promoter regions of different chordate and
plant genes
Endre Sebestyén†1, Tibor Nagy†2,3, Sándor Suhai3 and Endre Barta*2,3,4
Address: 1Agricultural Research Institute of the Hungarian Academy of Sciences, Martonvásár, Brunszvik u. 2., H-2462, Hungary, 2Bioinformatics
Group, Agricultural Biotechnology Center, Gödöllõ, Szent-Györgyi Albert u. 4., H-2100, Hungary, 3Department of Molecular Biophysics (B020),
German Cancer Research Center (DKFZ) Im Neuenheimer Feld 580, D-69120 Heidelberg, Germany and 4Apoptosis and Genomics Research
Group of the Hungarian Academy of Sciences, Research Center for Molecular Medicine, Medical and Health Science Center, University of
Debrecen, Debrecen, Hungary
Email: Endre Sebestyén - sebestyene@mail.mgki.hu; Tibor Nagy - black007@abc.hu; Sándor Suhai - s.suhai@dkfz-heidelberg.de;
Endre Barta* - barta@abc.hu
* Corresponding author †Equal contributors
Abstract
Background: The comparative genomic analysis of a large number of orthologous promoter regions of
the chordate and plant genes from the DoOP databases shows thousands of conserved motifs. Most of
these motifs differ from any known transcription factor binding site (TFBS). To identify common conserved
motifs, we need a specific tool to be able to search amongst them. Since conserved motifs from the DoOP
databases are linked to genes, the result of such a search can give a list of genes that are potentially regulated
by the same transcription factor(s).
Results: We have developed a new tool called DoOPSearch http://doopsearch.abc.hu for the analysis of
the conserved motifs in the promoter regions of chordate or plant genes. We used the orthologous
promoters of the DoOP database to extract thousands of conserved motifs from different taxonomic
groups. The advantage of this approach is that different sets of conserved motifs might be found depending
on how broad the taxonomic coverage of the underlying orthologous promoter sequence collection is
(consider e.g. primates vs. mammals or Brassicaceae vs. Viridiplantae). The DoOPSearch tool allows the
users to search these motif collections or the promoter regions of DoOP with user supplied query
sequences or any of the conserved motifs from the DoOP database. To find overrepresented gene
ontologies, the gene lists obtained can be analysed further using a modified version of the GeneMerge
program.
Conclusion: We present here a comparative genomics based promoter analysis tool. Our system is based
on a unique collection of conserved promoter motifs characteristic of different taxonomic groups. We offer
both a command line and a web-based tool for searching in these motif collections using user specified
queries. These can be either short promoter sequences or consensus sequences of known transcription
from European Molecular Biology Network (EMBnet) Conference 2008: 20th Anniversary Celebration
Martina Franca, Italy. 18–20 September 2008
Published: 16 June 2009
BMC Bioinformatics 2009, 10(Suppl 6):S6 doi:10.1186/1471-2105-10-S6-S6
<supplement> <title> <p>European Molecular Biology Network (EMBnet) Conference 2008: 20th Anniversary Celebration. Leading applications and technologies in bioinformatics</p> </title> <editor>Erik Bongcam-Rudloff, Domenica D'Elia, Andreas Gisel, Sophia Kossida, Kimmo Mattila and Lubos Klucar</editor> <note>Proceedings</note> <url>http://www.biomedcentral.com/content/pdf/1471-2105-10-S6-info.pdf</url> </supplement>
This article is available from: http://www.biomedcentral.com/1471-2105/10/S6/S6
© 2009 Sebestyén et al; licensee BioMed Central Ltd.
This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 9
(page number not for citation purposes)
factor binding sites. The GeneMerge analysis of the search results allows the user to identify statistically
overrepresented Gene Ontology terms that might provide a clue on the function of the motifs and genes.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
Background
The traditional bioinformatics approach for transcription
factor binding site (TFBS) discovery uses collections of
known (experimentally verified) TFBS sequences to find
their occurrences in a promoter. These sequences are usu-
ally described by a consensus sequence or a position spe-
cific matrix (PSM). There are several different databases,
whose main goal is to collect these consensus sequences
and matrices. Two of the better known collections are the
TRANSFAC [1] and JASPAR [2] databases. They contain
information for a large number of different TFBS groups.
Although the consensus sequence and matrix-based
approach has been used for almost twenty years [3], it still
has fundamental problems, such as having either too
many false positives or false negatives.
Another approach is to use sophisticated computer algo-
rithms designed for the ab initio prediction of overrepre-
sented sequence motifs in a collection of promoter
sequences [4,5]. In this case the user provides a number of
promoter regions, thought to be co-regulated or co-
expressed, and the computer algorithm identifies statisti-
cally overrepresented sequence oligos, putative TFBSs. As
the TFBSs are usually short, the promoter regions long,
and the bases can vary in certain positions, there are sim-
ilar problems, as in the method mentioned previously.
Besides the traditional analysis methods, the motif discov-
ery algorithms can also be used to find possible TFBSs in
the promoter regions of homologous gene sets. This
method is called phylogenetic footprinting [6], and an
important prerequisite of it is to collect as many ortholo-
gous promoter sequences as possible [7]. These sequences
are available in public DNA databases or in broadly
known genome browsers like ENSEMBL [8] or UCSC [9].
The most simple way to collect them however, is to use an
orthologous promoter database, such as OMGProm [10]
for the promoters of mammals, or DoOP [11] for the pro-
moters of chordates and plants. Recent developments in
comparative genomics allow us to search for conserved
motifs in the full genome of a large number of species. The
first collections of conserved sequence motifs are now
available for downloading and/or browsing. Xie et al. [12]
analysed the human, mouse, rat and dog whole genome
alignments, and described 174 sequence motifs that were
highly conserved. These motifs are now part of the JASPAR
database [2], thereby they can be used in a promoter anal-
ysis. The cisRED [13] database and the CORG [14] frame-
work are based on the human genome annotation from
ENSEMBL and on the COMPARA database respectively,
complemented by other sources. The extracted homolo-
gous gene sets in the cisRED database were analysed using
different motif discovery programs in order to find con-
motifs from the promoters of chordate and plant homol-
ogous gene sets. These motifs are derived from the con-
served regions of DIALIGN [16] alignments of the
orthologous promoters and thus represent a collinear set
of possible TFBSs.
There are a growing number of methods and websites that
offer promoter analysis tools and use different combina-
tions of the abovementioned approaches. In general, they
are more or less relying on the extra information coming
from comparative genomics methods, and orthologous
promoter collections [17-19].
Our aim was to develop a collection of web-based tools to
search with known or de novo discovered TFBSs, as well as
with longer promoter sequences, in order to collect and
examine in detail genes with similar conserved motifs in
their promoter regions. In this paper we describe the
DoOPSearch tool that provides a new and unique method
to search the motif collections of the DoOP database with
a new program called MOFEXT. The server also provides a
simple pattern search, based on FUZZNUC, in the pro-
moter regions of chordate and plant genes. To analyse the
Gene Ontology terms associated with a given gene in the
results, we integrated the GeneMerge program into DoOP-
Search. In addition, a Perl API called Bio::DOOP was also
developed for the easy manipulation and querying of the
DoOP database content.
Methods
Determining conserved motifs from orthologous promoters
of different taxonomic groups
In the original DoOP database [11], we used all the avail-
able orthologous promoter sequences of each gene to gen-
erate the multiple alignments and to determine the
evolutionary conserved motifs. Now we are using four
and ten different taxonomic categories in the chordate
and plant collections when selecting sequences for the
multiple alignments (Figure 1). The narrowest category
amongst the plants is the Brassicaceae family (Table 1),
which includes the Brassica species besides Arabidopsis
thaliana (as the originating species of the orthologous
clustering process). The following is the eudicotyledons
category, which also includes additional dicot species like
Medicago truncatula, Solanum lycopersicum, Populus tri-
chocarpa, Ricinus communis and others. The next one is the
Magnoliophyta, which includes the monocot species like
Oryza sativa and Zea mays, and finally the Viridiplantae,
which incorporates every species involved in the analysis.
We have defined ten taxonomic categories in the chordate
section (Table 2). The first one is the Primates, which con-
tains the Homo sapiens as the originating species and all
the monkeys. The following is the Euarchontoglires, whichPage 2 of 9
(page number not for citation purposes)
served sequence regions. The DoOP database [11,15] con-
tains more than one million different conserved sequence
contains the Primates and rodent species like the mouse
and rat. The next category is the Eutheria with all the pla-

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
Page 3 of 9
(page number not for citation purposes)
Different number of conserved motifs from different taxonomic groupsFigu 1
Different number of conserved motifs from different taxonomic groups. The PTPN23 (protein tyrosine phos-
phatase, non-receptor type 23) promoter cluster contains sequences from 19 different species. If the multiple alignment is
made from all the sequences (subset F) we only find one conserved motif (m1). If we narrow down the taxonomic group to
Theria (subset T), Eutheria (subset E) or Primates (subset P), we find 7, 9 and 23 conserved motifs respectively. The screen-
shots have been taken from the 500 base pair promoter cluster pages of the PTPN23 gene in the DoOP database.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
cental mammals. The Theria category also includes the
marsupials like the opossum, and the mammalian cate-
gory contains all the mammals including the egg-laying
platypus. The subsequent categories are the Amniota
(mammals as well as birds and reptiles), the Tetrapoda
(Amniota together with the amphibians like the Xenopus
species). The Teleostomi category also covers the fishes like
the fugu, zebrafish or Tetraodon nigroviridis. We also have
a category for the Vertebrate and Chordate taxonomical
groups, but due to the drawbacks of our orthologous find-
ing algorithm [11], they contain only a small number of
conserved motifs and are practically unusable.
Searching in the conserved motif collections and promoter
sequences
As a result of the orthologous promoter analysis, several
motif collections were defined, which consist of 6–50
base pair long consensus sequences. We developed a sim-
ple new program called MOFEXT (MOtiF sEarch and
EXTension) to perform fast gapless searches in the motif
collections containing even more than a million of con-
sensus sequences.
As input, the MOFEXT program requires one or more
query sequences and one or more motif lists. The program
slides a given length of window specified by the user, over
the query sequence, and starts to compare it with the first
consensus motif from the motif collections. Sliding the
window over the consensus motif, the program calculates
a score for each position using a predefined scoring
matrix. The overall score is then calculated by adding the
individual scores of each position together. The program
also calculates the maximum possible score for each win-
dow of the query sequence by comparing it with itself.
Using these scores, the program calculates a percentage
value for each pair of query and motif sequence window.
If this percentage value is larger than the cut-off value, and
the length of the query and the consensus motif from the
list is longer than the window size, the program tries to
extend the match in both directions. This means that the
program calculates the score as above for each possible
overlapping window with an increasing size between the
query and consensus sequence, and keeps the one with
the highest score. The MOFEXT program accepts the
standard IUPAC nucleotide.
For the search in the sequences of individual promoter
regions, the FUZZNUC program was used from the Euro-
pean Molecular Biology Open Software Suite (EMBOSS)
[20].
Developing a Perl API for handling and analysing the
DoOP data
In order to allow simple access of the data available in the
DoOP database we have developed a set of Perl modules
called Bio::DOOP. These modules are following the style
of the BIOPERL modules [21], and used mainly for query-
ing the MySQL backend of the DoOP database. The
Bio::DOOP modules are available from the CPAN Perl
repository http://search.cpan.org/~tibi/ with a complete
documentation. They are suitable for using in individual
PERL scripts for querying the DoOP database through the
net, and for building more complex promoter analysis
tools. It is possible for example to simply download spe-
cies-specific promoter sequences, determine the positions
of conserved motifs, or to download the conserved motif
sequences. The Bio::DOOP Perl modules also run and
parse the results of the MOFEXT program, the FUZZNUC
program from the EMBOSS package [20], and the Gene-
Merge Perl script [22].
Construction of the DoOPSearch tool
The DoOPSearch tool is running on a two processor Linux
server. The forms are generated using PHP scripts. The
programs are called and their results processed using Perl
scripts. Most of the data, including the search results, are
Table 2: Number of the different taxonomic groups in the
chordate database
Promoter size 500 1000 3000
Primates 21164 20484 18523
Euarchontoglires 14391 13665 12240
Eutheria 15411 14780 13744
Theria 4572 4390 4252
Mammalia 47 43 34
Amniota 1508 1373 1213
Tetrapoda 827 810 753
Teleostomi 1208 1161 1124
Vertebrata 9 2 1
Chordata 30 28 27
The data is taken from the DoOP database. The description of the
taxonomic groups is written in the text. The method for the
Table 1: Number of the different taxonomic groups in the plant
database
Promoter size 500 1000 3000
Brassicaceae 8207 4518 4257
eudicotyledons 2324 2263 2204
Magnoliophyta 370 277 267
Virdiplantae 41 27 21
The data is taken from the DoOP database. The description of the
taxonomic groups is written in the text. The method for the
generation of the taxonomic groups and conserved motifs is detailed
on the DoOP database web page. http://doop.abc.hu/creation.php.Page 4 of 9
(page number not for citation purposes)
stored in a MySQL database. The reading and writing of
the MySQL database is carried out with the Bio::DOOP
generation of the taxonomic groups and conserved motifs is detailed
on the DoOP database web page. http://doop.abc.hu/creation.php.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
MOFEXT or FUZZNUC search. The study size can be
altered by selecting and using different number of genes
on the MOFEXT or FUZZNUC search result page. Since
the results of the GeneMerge runs (especially the signifi-
cance of the evaluated GO terms) can vary considerably by
changing the initial MOFEXT and FUZZNUC search
parameters and/or the number of genes fed to the Gene-
Merge program, it is worth trying several runs and several
filtering parameters to find the best result.
Discussion
Searching for evolutionary conserved motifs
The first motif collections of the DoOP database were gen-
erated using all the available sequences of a given pro-
moter cluster. We noticed however, that using promoter
sequences from smaller taxonomic groups improves the
alignments significantly and thus can yield different con-
served motifs. The evolutionary sense of this approach is
the following. If for example a new TFBS had arisen in the
placental mammals (Eutheria), it would appear as con-
served motif only in an alignment made with only from
the Eutherian sequences. In the other alignments, which
in turn contain evolutionary more distant species, it can
be a non-conserved and so unnoticed region. It must be
also mentioned that different orthologous promoter clus-
ters can have different number of species, which in turn
can affect the determination of conserved motifs in our
algorithm. We defined the taxonomic categories based on
this logic. To our knowledge there are no similar motif
collections available, taking into account the evolutionary
distance between the species used in the phylogenetic
footprinting.
Case studies
Basically two types of analysis can be carried out with the
DoOPSearch tool. Either a longer promoter fragment, or a
short conserved motif, known TFBS can be used as a query
in the MOFEXT or FUZZNUC search.
First we demonstrate how a longer sequence can be used
for promoter analysis (Figure 2). Earlier we determined
both with experimental and in silico analysis tools, a pro-
moter element in the upstream region of the matrilin-1
gene [24]. In this example we used a 300 base pair
upstream fragment starting from the ATG start codon of
the human matrilin-1 gene available at the DoOP database
[15]. As a control, we chose the FABP4 gene and used the
same promoter region. After the MOFEXT search with
exactly the same parameters, we filtered the result and ran
the GeneMerge analysis. It is clear that there are specific
Gene Ontology categories overrepresented in each exam-
ple. It is also obvious that some categories like "transcrip-
tion" or "transcription factor activity" can be found in
promoter regions than other type of genes, but to confirm
this, we need to perform other analyses.
In the other case study (Figure 3) we used the binding site
of the NF-kappa B gene [25]. As a control we used the (not
reverse) complement of the NF-kappa B binding site. The
results clearly show the strength of our approach.
Although there are some significantly enriched GO terms
at the fake site as well, the difference proves the value of
our analysis.
These examples show how can the DoOPSearch tool help
the in silico annotation of longer promoter regions, known
TFBSs or conserved motifs with unknown functions. It
must be mentioned however, that to reach these result we
had to try different parameters both in the MOFEXT
search (word size, cut-off value and motif collection), and
also in the GeneMerge analysis.
Conclusion
The DoOPSearch tool in combination with the
Bio::DoOP Perl modules and the MOFEXT program are
suitable to search for conserved promoter motifs or
sequence patterns. Although there are a number of tools
and services available for computational analysis of tran-
scriptional regulation using different methods, to our
knowledge DoOPSearch is the first tool which gives the
opportunity to search in chordate or plant conserved
motif collections. To perform such a search, we have
developed a program called MOFEXT which utilizes a
simple gapless word matching and scoring algorithm in
order to search in a collections of conserved motifs, like
our motif collections from the DoOP database.
DoOPSearch is an ideal bioinformatics tool for research-
ers looking for potentially co-expressed genes, or putative
TFBSs in the upstream regions of different genes. Using
the MOFEXT search it is relatively easy to find genes con-
taining conserved motifs in their promoter regions similar
to the query. The query can either be a short experimen-
tally verified binding site or a promoter region with
unknown TFBSs but proven regulatory functions. By rec-
ognizing that the results from the MOFEXT or FUZZNUC
searches might contain co-expressed genes similar to the
results of microarray experiments, we offer a unique tool
to assign Gene Ontology terms and functions to a motif.
The method cerainly contains a lot of ambiguity and
might give false positive results, but after a detailed analy-
sis it might be a good indicator of the role of a TFBS in
gene regulation. We believe that using improved multiple
alignments (with more species for example) and using
more accurately defined promoter regions, our method
could be improved significantly and help in the in silicoPage 6 of 9
(page number not for citation purposes)
both results. The explanation for this can be that the tran-
scription factors contain more conserved motifs in their
annotation of TFBSs.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
Page 7 of 9
(page number not for citation purposes)
MOFEXT and GeneMerge analysis of the 300 base pair upstream region of the matrilin-1 and the FABP4 genesFigure 2
MOFEXT and GeneMerge analysis of the 300 base pair upstream region of the matrilin-1 and the FABP4 genes.
We downloaded the 500 base pair promoter region of the matrilin-1 (A1) and FABP4 (B1) genes. We used the last 300 base pair
of these sequences as a query in the MOFEXT search with the following parameters: wordsize: 8, cutoff: 70 and the 1000 base
pair E subset (A2 and B2). After the MOFEXT search we got 30548 (MATN1) and 23463 (FABP4) hits. We used the score range
151-40 (MATN1) and 105-40 (FABP4) for the GeneMerge analysis (A3 and B3). The genes in the GO term "Extracellular matrix
(sensu metazoan)" are listed in the panel A4. Some genes in the GO term "positive regulation of transcription from RNA
polymerase II promoter" are listed in the panel B4.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
Page 8 of 9
(page number not for citation purposes)
MOFEXT and GeneMerge analysis of the NF-kappa B binding siteFigure 3
MOFEXT and GeneMerge analysis of the NF-kappa B binding site. On the left side (1) we used the NF-kappa B binding
site consensus (GGGRNTTTCC, where R is A or G, and N is any base). On the right side we used the exact complement of
the previous site (CCCYNAAAGG, where Y is C or T). We used the same parameters for the MOFEXT search in both cases:
wordsize: 7, cutoff: 70 and the subset "All promoters, subset E" (A2 and B2). After the MOFEXT search we got 10697 (NF-
kappa B) and 9303 (fake site) hits. We used the score range 39-25 in both cases for the GeneMerge analysis. At the NF-kappa
B site, the genes from the GO category "lymph node developments" are listed.

BMC Bioinformatics 2009, 10(Suppl 6):S6 http://www.biomedcentral.com/1471-2105/10/S6/S6
We constantly develop and update the data and the web
interface. Due to the growing number of genome annota-
tions and available genome sequences, the quality of the
conserved motif collections will improve considerably.
We also: (I) try to refine and improve the existing motif
collections; (II) define and use new conserved motif col-
lections; (III) implement new filtering functions, allowing
the user to define a subset of genes (for example co-regu-
lated genes from microarray experiments) used in the
search; (IV) transfer the server to faster hardware and
improve the speed of the MOFEXT search by preliminary
runs, comparing the available motifs with each other
using default parameters.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
ES generated the motif subsets, modified the GeneMerge
script, designed the website and helped drafting the man-
uscript. TN wrote the MOFEXT program and the
Bio::DOOP Perl modules. EB initiated and coordinated
the project and wrote the manuscript. ES, TN, EB and SS
were all involved in the discussion of the data analysis and
assisted in writing the manuscript. All authors read and
approved the final manuscript.
Acknowledgements
This research was supported by grants from the Hungarian Scientific
Research Fund (OTKA) (NK60352 and NK72730) and from the Hungarian
Ministry of Agriculture and Rural Developmen R&D grant. E. Barta was sup-
ported by the Bolyai János Scholarship. We thank Dr Gábor Tóth and
Tamás Pálfy for their contribution in developing the original DoOP data-
bases and for providing their scripts for using in this work.
This article has been published as part of BMC Bioinformatics Volume 10 Sup-
plement 6, 2009: European Molecular Biology Network (EMBnet) Confer-
ence 2008: 20th Anniversary Celebration. Leading applications and
technologies in bioinformatics. The full contents of the supplement are
available online at http://www.biomedcentral.com/1471-2105/10?issue=S6.
References
1. Wingender E: The TRANSFAC project as an example of
framework technology that supports the analysis of genomic
regulation. Brief Bioinform 2008, 9:326-332.
2. Bryne JC, Valen E, Tang MH, Marstrand T, Winther O, da Piedade I,
Krogh A, Lenhard B, Sandelin A: JASPAR, the open access data-
base of transcription factor-binding profiles: new content
and tools in the 2008 update. Nucleic Acids Res 2008,
36:D102-106.
3. Bucher P: Weight matrix descriptions of four eukaryotic RNA
polymerase II promoter elements derived from 502 unre-
lated promoter sequences. J Mol Biol 1990, 212:563-578.
4. Tompa M, Li N, Bailey TL, Church GM, De Moor B, Eskin E, Favorov
AV, Frith MC, Fu Y, Kent WJ, et al.: Assessing computational
tools for the discovery of transcription factor binding sites.
Nat Biotechnol 2005, 23:137-144.
5. Rombauts S, Florquin K, Lescot M, Marchal K, Rouze P, Peer Y van
de: Computational approaches to identify promoters and cis-
regulatory elements in plant genomes. Plant Physiol 2003,
6. Blanchette M, Tompa M: Discovery of regulatory elements by a
computational method for phylogenetic footprinting.
Genome Res 2002, 12:739-748.
7. Thomas JW, Touchman JW, Blakesley RW, Bouffard GG, Beckstrom-
Sternberg SM, Margulies EH, Blanchette M, Siepel AC, Thomas PJ,
McDowell JC, et al.: Comparative analyses of multi-species
sequences from targeted genomic regions. Nature 2003,
424:788-793.
8. Flicek P, Aken BL, Beal K, Ballester B, Caccamo M, Chen Y, Clarke L,
Coates G, Cunningham F, Cutts T, et al.: Ensembl 2008. Nucleic
Acids Res 2008, 36:D707-714.
9. Karolchik D, Kuhn RM, Baertsch R, Barber GP, Clawson H, Diekhans
M, Giardine B, Harte RA, Hinrichs AS, Hsu F, et al.: The UCSC
Genome Browser Database: 2008 update. Nucleic Acids Res
2008, 36:D773-779.
10. Palaniswamy SK, Jin VX, Sun H, Davuluri RV: OMGProm: a data-
base of orthologous mammalian gene promoters. Bioinformat-
ics 2005, 21:835-836.
11. Barta E, Sebestyen E, Palfy TB, Toth G, Ortutay CP, Patthy L: DoOP:
Databases of Orthologous Promoters, collections of clusters
of orthologous upstream sequences from chordates and
plants. Nucleic Acids Res 2005, 33:D86-90.
12. Xie X, Lu J, Kulbokas EJ, Golub TR, Mootha V, Lindblad-Toh K,
Lander ES, Kellis M: Systematic discovery of regulatory motifs
in human promoters and 3' UTRs by comparison of several
mammals. Nature 2005, 434:338-345.
13. Robertson G, Bilenky M, Lin K, He A, Yuen W, Dagpinar M, Varhol
R, Teague K, Griffith OL, Zhang X, et al.: cisRED: a database sys-
tem for genome-scale computational discovery of regula-
tory elements. Nucleic Acids Res 2006, 34:D68-73.
14. Dieterich C, Grossmann S, Tanzer A, Ropcke S, Arndt PF, Stadler PF,
Vingron M: Comparative promoter region analysis powered
by CORG. BMC Genomics 2005, 6:24.
15. Barta E: Comparative genomics-based orthologous promoter
analysis using the DoOP database and the DoOPSearch web
tool. Methods Mol Biol 2007, 395:319-328.
16. Morgenstern B: DIALIGN 2: improvement of the segment-to-
segment approach to multiple sequence alignment. Bioinfor-
matics 1999, 15:211-218.
17. Ward LD, Bussemaker HJ: Predicting functional transcription
factor binding through alignment-free and affinity-based
analysis of orthologous promoter sequences. Bioinformatics
2008, 24:i165-171.
18. Hooghe B, Hulpiau P, van Roy F, De Bleser P: ConTra: a promoter
alignment analysis tool for identification of transcription fac-
tor binding sites across species. Nucleic Acids Res 2008,
36:W128-132.
19. Gotea V, Ovcharenko I: DiRE: identifying distant regulatory ele-
ments of co-expressed genes. Nucleic Acids Res 2008,
36:W133-139.
20. Rice P, Longden I, Bleasby A: EMBOSS: the European Molecular
Biology Open Software Suite. Trends Genet 2000, 16:276-277.
21. Stajich JE, Block D, Boulez K, Brenner SE, Chervitz SA, Dagdigian C,
Fuellen G, Gilbert JG, Korf I, Lapp H, et al.: The Bioperl toolkit:
Perl modules for the life sciences. Genome Res 2002,
12:1611-1618.
22. Castillo-Davis CI, Hartl DL: GeneMerge – post-genomic analy-
sis, data mining, and hypothesis testing. Bioinformatics 2003,
19:891-892.
23. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM,
Davis AP, Dolinski K, Dwight SS, Eppig JT, et al.: Gene ontology:
tool for the unification of biology. The Gene Ontology Con-
sortium. Nat Genet 2000, 25:25-29.
24. Rentsendorj O, Nagy A, Sinko I, Daraba A, Barta E, Kiss I: Highly
conserved proximal promoter element harbouring paired
Sox9-binding sites contributes to the tissue- and develop-
mental stage-specific activity of the matrilin-1 gene. Biochem
J 2005, 389:705-716.
25. Kunsch C, Ruben SM, Rosen CA: Selection of optimal kappa B/
Rel DNA-binding motifs: interaction of both subunits of NF-
kappa B with DNA is required for transcriptional activation.
Mol Cell Biol 1992, 12:4412-4421.Page 9 of 9
(page number not for citation purposes)
132:1162-1176.