BIPS: BIANA Interolog Prediction Server. A tool for protein-protein interaction inference.
ABSTRACT Protein-protein interactions (PPIs) play a crucial role in biology, and high-throughput experiments have greatly increased the coverage of known interactions. Still, identification of complete inter- and intraspecies interactomes is far from being complete. Experimental data can be complemented by the prediction of PPIs within an organism or between two organisms based on the known interactions of the orthologous genes of other organisms (interologs). Here, we present the BIANA (Biologic Interactions and Network Analysis) Interolog Prediction Server (BIPS), which offers a web-based interface to facilitate PPI predictions based on interolog information. BIPS benefits from the capabilities of the framework BIANA to integrate the several PPI-related databases. Additional metadata can be used to improve the reliability of the predicted interactions. Sensitivity and specificity of the server have been calculated using known PPIs from different interactomes using a leave-one-out approach. The specificity is between 72 and 98%, whereas sensitivity varies between 1 and 59%, depending on the sequence identity cut-off used to calculate similarities between sequences. BIPS is freely accessible at http://sbi.imim.es/BIPS.php.
- [Show abstract] [Hide abstract]
ABSTRACT: Protein-protein interactions play a relevant role among the different functions of a cell. Identifying the protein-protein interaction network of a given organism (interactome) is useful to shed light on the key molecular mechanisms within a biological system. In this work, we show the role of structural features (loops and domains) to comprehend the molecular mechanisms of protein-protein interactions. A paradox in protein-protein binding is to explain how the unbound proteins of a binary complex recognize each other among a large population within a cell and how they find their best docking interface in a short time-scale. We use interacting and non-interacting protein pairs to classify the structural features that sustain the binding (or non-binding) behaviour. Our study indicates that not only the interacting region but also the rest of the protein surface is important for the interaction fate. The interpretation of this classification suggests that the balance between favouring and disfavouring structural features determines if a pair of proteins interacts or not. Our results are in agreement with previous works and support the funnel-like intermolecular energy landscape theory that explains protein-protein interactions. We have used these features to score the likelihood of the interaction between two proteins and to develop a method for the prediction of PPIs. We have tested our method on several sets with unbalanced ratios of interactions and non-interactions to simulate real conditions, obtaining accuracies higher than 25% in the most unfavourable circumstances.Journal of Molecular Biology 01/2013; · 3.91 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Protein interaction networks are important for the understanding of regulatory mechanisms, for the explanation of experimental data and for the prediction of protein functions. Unfortunately, most interaction data is available only for model organisms. As a possible remedy, the transfer of interactions to organisms of interest is common practice, but it is not clear when interactions can be transferred from one organism to another and, thus, the confidence in the derived interactions is low. Here, we propose to use a rich set of features to train Random Forests in order to score transferred interactions. We evaluated the transfer from a range of eukaryotic organisms to S. cerevisiae using orthologs. Directly transferred interactions to S. cerevisiae are on average only 24% consistent with the current S. cerevisiae interaction network. By using commonly applied filter approaches the transfer precision can be improved, but at the cost of a large decrease in the number of transferred interactions. Our Random Forest approach uses various features derived from both the target and the source network as well as the ortholog annotations to assign confidence values to transferred interactions. Thereby, we could increase the average transfer consistency to 85%, while still transferring almost 70% of all correctly transferable interactions. We tested our approach for the transfer of interactions to other species and showed that our approach outperforms competing methods for the transfer of interactions to species where no experimental knowledge is available. Finally, we applied our predictor to score transferred interactions to 83 targets species and we were able to extend the available interactome of B. taurus, M. musculus and G. gallus with over 40,000 interactions each. Our transferred interaction networks are publicly available via our web interface, which allows to inspect and download transferred interaction sets of different sizes, for various species, and at specified expected precision levels. Availability:http://services.bio.ifi.lmu.de/coin-db/.PLoS ONE 01/2013; 8(6):e66635. · 3.53 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Protein interaction maps are the key to understand the complex world of biological processes inside the cell. Public protein databases have already catalogued hundreds of thousands of experimentally discovered interactions, and struggle to curate all the existing information dispersed through the literature. However, to be most efficient, standard protocols need to be implemented for direct submission of new interaction sets directly into databases. At the same time, great efforts are invested to expand the coverage of the interaction space and unveil the molecular details of such interactions up to the atomistic level. The net result will be the definition of a detailed atlas spanning the universe of protein interactions to guide the everyday work of the biologist.Current Opinion in Structural Biology 07/2013; 23(6):929-940. · 8.74 Impact Factor
BIPS: BIANA Interolog Prediction Server. A tool for
protein–protein interaction inference
Javier Garcia-Garcia1, Sylvia Schleker2, Judith Klein-Seetharaman2,3and Baldo Oliva1,*
1Structural Bioinformatics Laboratory (GRIB-IMIM), Universitat Pompeu Fabra, Barcelona Research Park of
Biomedicine (PRBB), 08003 Barcelona, Catalonia, Spain,2Forschungszentrum Ju ¨lich, Institute of Complex
Systems (ICS-5), 52425 Ju ¨lich, Germany and3Department of Structural Biology, University of Pittsburgh,
Pittsburgh, PA 15260, USA
Received February 25, 2012; Revised May 16, 2012; Accepted May 17, 2012
Protein–protein interactions (PPIs) play a crucial role
in biology, and high-throughput experiments have
greatly increased the coverage of known inter-
actions. Still, identification of complete inter- and
complete. Experimental data can be complemented
by the prediction of PPIs within an organism or
between two organisms based on the known
interactions of the orthologous genes of other
BIANA (Biologic Interactions and Network Analysis)
Interolog Prediction Server (BIPS), which offers a
web-based interface to facilitate PPI predictions
based on interolog information. BIPS benefits from
the capabilities of the framework BIANA to integrate
metadata can be used to improve the reliability of
the predicted interactions. Sensitivity and specificity
of the server have been calculated using known
PPIs from different interactomes using a leave-
72 and 98%, whereas sensitivity varies between 1
and 59%, depending on the sequence identity
cut-off used tocalculate
The number of protein–protein interactions (PPIs) experi-
mentally obtained for proteomes of entire species is
smaller than the number of expected PPIs (1). Even the
number of interactions between proteins in yeast, one of
the most studied organisms, is still believed to be
underestimated (2). To complement experimental data,
several computational methods have been developed to
predict PPIs (3,4). One example is the use of experimen-
tally identified interactions in one organism to predict
the interactions in other organisms assuming that
homolog proteins preserve their ability to interact (5,6).
The basis of this hypothesis is to assume that homologs
have similar functional behaviour; therefore, they preserve
the same PPIs. The prediction of the interaction between
the homologs of two interacting proteins is defined as
interolog (conservation of PPIs). Several works have
used interologsto extend
interactomes (7,8) or to predict pathogen/host cross-
species PPIs (9,10). Up to now, most predictions based
on interologs are restricted to a few species or a limited
set of template interactions. For example, Yu et al. (5)
transferred the known interactions of yeast to four
different species. Similarly, Wiles et al. (11) developed
InterologFinder, a tool to map the interactions integrated
in MiMi (12) to five species. Interestingly, PPISearch (13)
implements the interolog approach providing different
scoring functions, but is restricted to the analysis of a
single protein pair at any submission instance. Recently,
Gallone et al. (14) developed a Perl module to automate
predictions based on interologs, using optional metadata
to rank the interactions. However, this still requires
programming skills. The current limitations stem from
the fact that information of known PPIs is spread
among several repositories, and sets of PPIs from different
databases show a low intersection (15,16). This challenge
has led to the development of data integration strategies,
such as Biologic Interactions and Network Analysis
(BIANA) (17). BIANA is a program framework used in
the integration of biological databases mostly focused on
Here, we present the BIANA Interolog Prediction
Server (BIPS). BIPS offers a web interface to facilitate
the prediction of PPIs based on interologs for a set of
proteins provided by the user as input, including entire
proteomes. The server benefits from the integration
capabilities of BIANA to use a large data set of experi-
mentally identified PPIs. BIANA also offers additional
*To whom correspondence should be addressed. Tel: +34 93 316 05 09; Fax: +34 93 316 05 50; Email: email@example.com
Published online 11 June 2012Nucleic Acids Research, 2012, Vol. 40, Web Server issue W147–W151
? The Author(s) 2012. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/
by-nc/3.0), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
information such as gene ontology (GO) terms, clusters of
orthologous genes and many other attributes such as pre-
dicted number of transmembrane helices, which gives the
user the freedom to restrict the predictions according to
MATERIALS AND METHODS
The interolog hypothesis implies that two proteins (A and
B) are predicted to interact if a known interaction between
two proteins (A’ and B’) exists, such that A is similar to A’
and B similar to B’. The interaction between the proteins
A and B is called target interaction (with A and B being
defined as protein targets), whereas the interaction
between proteins A’ and B’ is called template interaction
(with A’ and B’ being defined as templates). In the BIPS
server, protein A is the query protein submitted by the
user, and protein B is the predicted partner. This broad
definition of interologs implies that the hypothesis works
not only for orthologs but also for paralogs of the
Sequence similarity measure
Sequence similarity between proteins relies on basic local
alignment search tool (BLAST) alignments (18). Query
protein sequences are aligned against all sequences with
known interactions stored in the BIANA MySQL
database (17). The alignments provide a similarity
measure based on the percentage of identical residues
aligned and the percentage of the sequence length of the
queries and templates covered by the alignment (query and
template coverage, respectively). We use a threshold of
90% of template coverage to ensure that the prediction
is not inferred from local regions of the template inter-
action. The geometric mean of individual identities (joint
identities) and the geometric mean of individual BLAST
E-values (joint E-value) are also considered, as proposed
by Yu et al. (5). The BIPS server uses a local database of
stored similarity measures to avoid unnecessary repeated
BLAST searches. This speeds up the server, allowing users
to obtain predictions of interactions of full proteomes in
manageable time. In this manner, only entirely new se-
quences consume extra time in the first run.
We hypothesize that protein A interacts with protein B
if a domain A’ can be assigned to A and a domain B’ to
B, such that A’ and B’ are interacting domains in the
iPfam (19) or the 3DID (20) database. We measure the
similarity of the target sequences (A and B) with Pfam
domains as a function of the E-value calculated with the
package HMMER 3.0 (21). We assign Pfam domains
using an E-value cut-off of 10?5in the Pfam A database.
Template interactions were extracted using the BIANA
framework (17) integrating the following 10 databases:
DIP (22), HPRD (23), IntAct (24), MINT (25), MPact
(26), PHI_base (27), PIG (28), BioGRID (29), BIND
(30) and VirusMINT (31). It has been noted that PPI
databases share little overlap (16). Therefore, using the
integration of multiple sources instead of a single source
greatly enlarges the set of predictions. Furthermore, we
have used BIANA to include information from other
databases such as Uniprot (32) and GO (33). This add-
itional information can be used to interpret the prediction
results and select predicted interactions deemed interesting
e.g. for experimental validation. Finally, sources of
domain–domain interactions are also included, using
iPfam (19) and 3DID (20).
Interacting proteins likely share biological processes or
share similar locations compared with non-interacting
proteins (4). Thus, we can use a number of similar func-
tional annotations between each pair of proteins predicted
to interact to rank the predictions. BIPS uses GO
annotations to select the most probable prediction for a
query protein by selecting those partners that share similar
GO terms. The similarity between GO terms implies that
either they are equal or there is a parenthood relationship
between them. In addition, GO term semantic similarity
and the functional similarity of genes are computed
using the method proposed by Wang et al. (34)
Clusters of orthologous genes
Two proteins are considered orthologous if they are
included in the same cluster of orthologous genes.
Ortholog definitions between proteins are extracted from
eggNOG (35) and Ensembl Compara (36) databases. Our
predictions can be filtered assuming the traditional defin-
ition of interologs: two target proteins are supposed to
interact if they are orthologous to two known interaction
DESCRIPTION OF THE WEBSERVER
Proteins for which the user wants to predict putative
binding partners can be uploaded as a list of sequences
in FASTA format or a list of protein identifiers
Uniprot entry and gene
The output is a list of predictions that can be viewed or
downloaded. The user can browse the data associated with
the predicted partners. Some details of the template inter-
action, such as the source database of the interaction and
the method of detection, are provided. The user can select
several parameters helping estimate the reliability of the
predictions: (i) sequence similarity measures, (ii) checking
domain–domain interactions (either using domains from
3DID or iPfam), (iii) checking common GO terms
between the predicted partners of an interaction and
(iv) using clusters of orthologous genes for the prediction.
Additionally, template interactions can be restricted to
a subset of proteins to improve the reliability of the pre-
dictions. For example, the user can select interactions
W148Nucleic Acids Research, 2012,Vol.40, Web Server issue
based on the experimental methods by which the template
interactions were observed, the number of experiments
confirming the template interaction or the number of
species in which the interaction between homolog pairs
of the template was observed.
Finally, the user can restrict the list of predictions,
reducing the number of predictions to a manageable
size. The user can select specific partners: (i) those with
specific keywords in their descriptive attributes, (ii) those
associated with a certain pathology, (iii) those belonging
to a specific taxon, including the case in which query
proteins are from a particular pathogen, and the predicted
partners are from selected hosts, (iv) those belonging to
a subset of proteins uploaded by the user and (v) those
with transmembrane predicted regions [calculated with
We have checked the validity of our predictions by two
interactomes reported in BIANA were predicted using
the leave-one-out strategy. For each interaction being
tested, all interactions reported in BIANA were used as
templates including the interactions of the same organism,
but excluding the interaction being tested. Several organ-
isms covering different taxonomy groups were considered
(see Table 1). Sensitivity was calculated by testing the
percentage of known interactions correctly predicted
over the total of known interactions. Between 1 and
59% of known interactions were predicted by using a
sequence identity cut-off ranging from 30 to 90%. In
a complementary approach, we used the negatome
database (38) as a source for non-interactions to calculate
specificity. Specificity was calculated as the percentage of
correctly predicted negative interactions (true negatives)
out of 1291 non-interacting pairs. The specificity ranged
between 72 and 98%. However, the study of specificity is
not enough to quantify false positive predictions in the
PPI context because the number of non-interacting
protein pairs is larger than the number of interacting
pairs. Indeed, a high specificity does not necessarily
mean that a large proportion of the predicted interactions
is correct [i.e. positive predictive value (PPV)]. We selected
the Arabidopsis interactome (39) to estimate an example of
the PPV of our predictions (see Table 1). This is one of the
newest data sets of PPIs available, and we have to note
that its authors estimated that their experiment had
already a precision of 80% and a sensitivity of 16%. A
more detailed comparison with a previous interolog
approach is shown in Supplementary Table S1.
We have presented a PPI prediction server, BIPS, which is
based on the similarity between protein sequences and
PPIs reported in several biological databases integrated
using the BIANA framework. BIPS benefits from the
large amount of information deposited in these databases.
By increasing the number of template interactions,
coverage ofthe predictions
Traditionally, the interolog approach has been defined
as the transference of interactions between orthologs
from different species. However, the distinction between
orthologs (gene pairs that trace back to speciation) and
paralogs (gene pairs resulting from duplication events) is
not always clear, as it depends on the last common
ancestor applied. BIPS makes no distinction between
orthologs and paralogs. In contrast, BIPS relies directly
on pair-to-pair similarities to perform its inferences.
Comparison of predictions based only on groups of
between 30 and 70% identity (see Table 1). The main
advantages of BIPS over other servers are the capability
to predict interactions on a large scale such as for whole
proteomes, in a reasonable time, the use of up-to-date
database information and the option for the user to
select several filters to improve confidence in the results.
The latter is based on the notion that using additional
information about the protein targets and template inter-
actions increases the reliability of the predictions. The
sequence similarity measure is restrictive and some filters
Table 1.Sensitivity (a), specificity (b) and PPV (c) of the prediction of different data sets by varying the filtering conditions
Data set90% identity 70% identity30% identity eggNOG
All DomGO COGAll Dom GO COGAllDom GO COGAll
The percentage indicates sensitivity (a), specificity (b) or PPV (c). (d) Main screen data set from the Arabidopsis interactome map for 8596
Arabidopsis proteins, with a precision of 80% and a sensitivity of 16% (38). All, all predictions without applying any restriction; Dom, predictions
filtered by known interacting domains reported in 3DID or iPfam; GO, predictions filtered by GO term similarity (biological process or cellular
compartment); COG, predictions filtered by known interacting proteins in the same clusters of orthologous groups. Clusters of orthologous genes
were as defined in the eggNOG database excluding non-supervised clusters.
Nucleic AcidsResearch, 2012, Vol.40, WebServer issue W149
are applied (see sensitivity and specificity values in
Table 1). All these capabilities provide the user with a
useful tool to select predictions and focus on a specific
Supplementary Data are available at NAR Online:
Supplementary Table 1.
Spanish Ministry of Science and Innovation (MICINN);
German Federal Ministry of Education and Research
(BMBF); partners of the ERASysBio+ initiative sup-
ported under the EU ERA-NET Plus scheme in FP7
(SHIPREC)Euroinvestigacio ´ n
well as FEDER [BIO2011-22568, PSE-0100000-2009].
Funding for the open access charge: ERASysBio+initia-
tive supported under the EU ERA-NET Plus schme
in FP7 (SHIPREC) Euroinvestigacio ´ n [EUI2009-04018].
Conflict of interest statement. None declared.
1. Aloy,P. and Russell,R.B. (2004) Ten thousand interactions for the
molecular biologist. Nat. Biotechnol., 22, 1317–1321.
2. Sambourg,L. and Thierry-Mieg,N. (2010) New insights into
protein-protein interaction data lead to increased estimates of the
S. cerevisiae interactome size. BMC Bioinformatics, 11, 605.
3. Qi,Y., Klein-Seetharaman,J. and Bar-Joseph,Z. (2005) Random
forest similarity for protein-protein interaction prediction from
multiple sources. Pac. Symp. Biocomput., 531–542.
4. Jain,S. and Bader,G.D. (2010) An improved method for scoring
protein-protein interactions using semantic similarity within the
gene ontology. BMC Bioinformatics, 11, 562.
5. Yu,H., Luscombe,N.M., Lu,H.X., Zhu,X., Xia,Y., Han,J.D.,
Bertin,N., Chung,S., Vidal,M. and Gerstein,M. (2004) Annotation
transfer between genomes: protein-protein interologs and
protein-DNA regulogs. Genome Res., 14, 1107–1118.
6. Matthews,L.R., Vaglio,P., Reboul,J., Ge,H., Davis,B.P.,
Garrels,J., Vincent,S. and Vidal,M. (2001) Identification of
potential interaction networks using sequence-based searches for
conserved protein-protein interactions or ‘‘interologs’’. Genome
Res., 11, 2120–2126.
7. Wang,F., Liu,M., Song,B., Li,D., Pei,H., Guo,Y., Huang,J. and
Zhang,D. (2012) Prediction and characterization of protein-
protein interaction networks in swine. Proteome Sci., 10, 2.
8. Shin,C.J., Davis,M.J. and Ragan,M.A. (2009) Towards the
mammalian interactome: inference of a core mammalian
interaction set in mouse. Proteomics, 9, 5256–5266.
9. Schleker,S., Garcia-Garcia,J., Klein-Seetharaman,J. and Oliva,B.
(2012) Prediction and comparison of Salmonella-human and
Salmonella-Arabidopsis interactomes. Chem Biodivers, 9, 991–1018.
10. Krishnadev,O. and Srinivasan,N. (2011) Prediction of
protein-protein interactions between human host and a pathogen
and its application to three pathogenic bacteria. Int. J. Biol
Macromol., 48, 613–619.
11. Wiles,A.M., Doderer,M., Ruan,J., Gu,T.T., Ravi,D., Blackman,B.
and Bishop,A.J. (2010) Building and analyzing protein
interactome networks by cross-species comparisons. BMC Syst.
Biol., 4, 36.
12. Tarcea,V.G., Weymouth,T., Ade,A., Bookvich,A., Gao,J.,
Mahavisno,V., Wright,Z., Chapman,A., Jayapandian,M.,
Ozgur,A. et al. (2009) Michigan molecular interactions r2: from
interacting proteins to pathways. Nucleic Acids Res., 37,
13. Chen,C.C., Lin,C.Y., Lo,Y.S. and Yang,J.M. (2009) PPISearch:
a web server for searching homologous protein-protein
interactions across multiple species. Nucleic Acids Res., 37,
14. Gallone,G., Simpson,T.I., Armstrong,J.D. and Jarman,A.P. (2011)
Bio::Homology::InterologWalk—a Perl module to build putative
protein-protein interaction networks through interolog mapping.
BMC Bioinformatics, 12, 289.
15. Mathivanan,S., Periaswamy,B., Gandhi,T.K., Kandasamy,K.,
Suresh,S., Mohmood,R., Ramachandra,Y.L. and Pandey,A.
(2006) An evaluation of human protein-protein interaction data in
the public domain. BMC Bioinformatics, 7 (Suppl. 5), S19.
16. Aragues,R., Garcia-Garcia,J. and Oliva,B. (2008) Integration and
prediction of PPI using multiple resources from public databases.
J. Proteomics. Bioinform., 1, 166–187.
17. Garcia-Garcia,J., Guney,E., Aragues,R., Planas-Iglesias,J. and
Oliva,B. (2010) Biana: a software framework for compiling
biological interactions and analyzing networks. BMC
Bioinformatics, 11, 56.
18. Altschul,S.F., Gish,W., Miller,W., Myers,E.W. and Lipman,D.J.
(1990) Basic local alignment search tool. J. Mol. Biol., 215,
19. Finn,R.D., Marshall,M. and Bateman,A. (2005) iPfam:
visualization of protein-protein interactions in PDB at domain
and amino acid resolutions. Bioinformatics, 21, 410–412.
20. Stein,A., Ceol,A. and Aloy,P. (2011) 3did: identification and
classification of domain-based interactions of known
three-dimensional structure. Nucleic Acids Res., 39, D718–D723.
21. Finn,R.D., Clements,J. and Eddy,S.R. (2011) HMMER web
server: interactive sequence similarity searching. Nucleic Acids
Res., 39, W29–W37.
22. Salwinski,L., Miller,C.S., Smith,A.J., Pettit,F.K., Bowie,J.U. and
Eisenberg,D. (2004) The Database of Interacting Proteins: 2004
update. Nucleic Acids Res., 32, D449–D451.
23. Keshava Prasad,T.S., Goel,R., Kandasamy,K., Keerthikumar,S.,
Kumar,S., Mathivanan,S., Telikicherla,D., Raju,R., Shafreen,B.,
Venugopal,A. et al. (2009) Human Protein Reference Database—
2009 update. Nucleic Acids Res., 37, D767–D772.
24. Kerrien,S., Alam-Faruque,Y., Aranda,B., Bancarz,I., Bridge,A.,
Derow,C., Dimmer,E., Feuermann,M., Friedrichsen,A.,
Huntley,R. et al. (2007) IntAct—open source resource for
molecular interaction data. Nucleic Acids Res., 35, D561–D565.
25. Licata,L., Briganti,L., Peluso,D., Perfetto,L., Iannuccelli,M.,
Galeota,E., Sacco,F., Palma,A., Nardozza,A.P., Santonico,E.
et al. (2011) MINT, the molecular interaction database: 2012
update. Nucleic Acids Res., 40, D857–D861.
26. Guldener,U., Munsterkotter,M., Oesterheld,M., Pagel,P.,
Ruepp,A., Mewes,H.W. and Stumpflen,V. (2006) MPact: the
MIPS protein interaction resource on yeast. Nucleic Acids Res.,
27. Winnenburg,R., Baldwin,T.K., Urban,M., Rawlings,C., Kohler,J.
and Hammond-Kosack,K.E. (2006) PHI-base: a new database for
pathogen host interactions. Nucleic Acids Res., 34, D459–D464.
28. Driscoll,T., Dyer,M.D., Murali,T.M. and Sobral,B.W. (2009)
PIG—the pathogen interaction gateway. Nucleic Acids Res., 37,
29. Stark,C., Breitkreutz,B.J., Chatr-Aryamontri,A., Boucher,L.,
Oughtred,R., Livstone,M.S., Nixon,J., Van Auken,K., Wang,X.,
Shi,X. et al. (2011) The BioGRID Interaction Database: 2011
update. Nucleic Acids Res., 39, D698–D704.
30. Bader,G.D., Betel,D. and Hogue,C.W. (2003) BIND: the
Biomolecular Interaction Network Database. Nucleic Acids Res.,
31. Chatr-aryamontri,A., Ceol,A., Peluso,D., Nardozza,A., Panni,S.,
Sacco,F., Tinti,M., Smolyar,A., Castagnoli,L., Vidal,M. et al.
(2009) VirusMINT: a viral protein interaction database. Nucleic
Acids Res., 37, D669–D673.
32. UniProt Consortium. (2012) Reorganizing the protein space at
the Universal Protein Resource (UniProt). Nucleic Acids Res., 40,
33. Ashburner,M., Ball,C.A., Blake,J.A., Botstein,D., Butler,H.,
Cherry,J.M., Davis,A.P., Dolinski,K., Dwight,S.S., Eppig,J.T.
et al. (2000) Gene ontology: tool for the unification of biology.
The Gene Ontology Consortium. Nat. Genet., 25, 25–29.
W150Nucleic Acids Research, 2012,Vol.40, Web Server issue
34. Wang,J.Z., Du,Z., Payattakool,R., Yu,P.S. and Chen,C.F. (2007)
A new method to measure the semantic similarity of GO terms.
Bioinformatics, 23, 1274–1281.
35. Powell,S., Szklarczyk,D., Trachana,K., Roth,A., Kuhn,M.,
Muller,J., Arnold,R., Rattei,T., Letunic,I., Doerks,T. et al. (2012)
eggNOG v3.0: orthologous groups covering 1133 organisms at
41 different taxonomic ranges. Nucleic Acids Res., 40,
36. Flicek,P., Amode,M.R., Barrell,D., Beal,K., Brent,S., Chen,Y.,
Clapham,P., Coates,G., Fairley,S., Fitzgerald,S. et al. (2011)
Ensembl 2011. Nucleic Acids Res., 39, D800–D806.
37. Sonnhammer,E.L., von Heijne,G. and Krogh,A. (1998) A hidden
Markov model for predicting transmembrane helices in protein
sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol., 6, 175–182.
38. Smialowski,P., Pagel,P., Wong,P., Brauner,B., Dunger,I.,
Fobo,G., Frishman,G., Montrone,C., Rattei,T., Frishman,D.
et al. (2010) The Negatome database: a reference set of
non-interacting protein pairs. Nucleic Acids Res., 38,
39. Arabidopsis Interactome Mapping Consortium. (2011) Evidence
for network evolution in an Arabidopsis interactome map.
Science, 333, 601–607.
Nucleic AcidsResearch, 2012, Vol.40, WebServer issueW151