iPath: interactive exploration of biochemical pathways
Ivica Letunic1, Takuji Yamada1, Minoru Kanehisa2and Peer Bork1
1EMBL, Meyerhofstrasse 1, 69117 Heidelberg, Germany
2Bioinformatics Center, Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
iPath is an open-access online tool (http://pathways.
embl.de) for visualizing and analyzing metabolic path-
ways. An interactive viewer provides straightforward
navigation through various pathways and enables easy
access to the underlying chemicals and enzymes. Cus-
tomized pathway maps can be generated and annotated
using various external data. For example, by merging
human genome data with two important gut commen-
sals, iPath can pinpoint the complementarity of the
host–symbiont metabolic capacities.
New era of pathway exploration
The recent publication of the KEGG (http://www.genome.
jp/kegg/) global overview map of metabolic pathways in
Scalable Vector Graphics format  represents an import-
ant step towards large-scale visualization and interpret-
ation of various data regarding metabolic activities. The
global map has been manually constructed using 123
classical KEGG maps with an average of 17 reactions each;
the result is an overview of a large proportion of all
metabolic reactions known to date, collected from various
biological systems (see Figure 1 for details).
Toaccess anddigest thisvastamountofinformation,we
have developed an interactive Pathways Explorer, iPath,
which provides powerful tools to visualize, navigate,
explore and analyze all, or a subset of, various pathway
maps. Pathway maps in iPath are accessed through an
interactive online viewer, which provides simple zooming
and panning controls, with different levels of map detail
corresponding to various zoom levels. Detailed information
is available for various nodes and edges (lines) in the map,
such as for the enzymes, reactions and compounds
involved. Further information can be accessed via hyper-
links to other online resources, for example, KEGG ,
(http://www.3dmet.dna.affrc.go.jp/) and ChEBI .
iPath can export the maps into various graphical for-
mats, which can be further modified or included in various
documents and publications.
Mapping data onto pathways and customizing maps
iPath provides a simple way of creating customized path-
way mapsthroughdata mapping. Severaltypes of datacan
be used to change every part of the map, such as pathway
and compound identifiers, protein accession numbers,
COG , eggNOG , KEGG orthologous group identifiers
and enzyme EC numbers. In addition, colors, opacity and
width can be specified for any node or edge in the map.
Queries can also be performed using enzyme names,
enabling easy highlighting of various pathways where
individual or groups of enzymes occur.
These features create a powerful framework for the
exploration and analysis of particular metabolic pathways
or overall metabolism, and for comparative analyses of
various genomics, transcriptomics or proteomics datasets.
by iPath include overviews of the metabolic capacity
encoded by a single (meta)genome, exploration of meta-
bolic differences in various spatial and temporal series
datasets, comparative and evolutionary studies with many
organisms in addition to species-complementarity studies.
The latter is illustrated by a straightforward analysis to
identify pathways in abundant human gut bacteria that
complement those in human (Figure 1).
Figure 1 shows all known, categorized metabolic path-
ways of Homo sapiens and the complementary bacterial
pathways of Bifidobacterium longum and Methanobrevi-
bacter smithii. These two bacteria inhabit the human
intestine, and are an integral part of our intestinal micro-
flora [5–7]. Among the many pathways that are specific to
B. longum and M. smithii, are those of peptidoglycan and
tantly, B. longum and M. smithii synthesize five com-
pounds that are missing in human metabolism, namely,
vitamins B1, B2, B5, B9 and H, which can all be absorbed
in the human large intestine . The respective com-
plementary pathways that are responsible for the syn-
thesis of these vitamins are clearly visible in iPath
(Figure 1). iPath further reveals that the two bacteria also
encode, in contrast to humans, the enzymes needed for the
synthesis of several other vitamins (such as B12, K2 and
B3); whether these vitamins can also be absorbed by the
human large intestine still needs to be determined.
It is well known that bacteria provide various effective
still unclear which compounds are synthesized by human
endosymbiotic bacteria. The simple analysis shown here
demonstrates the power of iPath in hypothesis generation,
comparative pathway analysis and many other appli-
Exploring the metabolism of various species
of species-specific metabolic pathways maps, which were
created using ortholog definitions for enzymatic proteins of
each of the 183 species present in the global map (21
Corresponding author: Bork, P. (email@example.com).