Biology Reference
In-Depth Information
reasoning, super rules and increasing rule stringency are introduced to meet the need of
users that are not looking for too many possible answers.
38
The UM-PPS is part of the
University of Minnesota Biocatalysis/Biodegradation Database (UM-BBD), a database
dedicated to information regarding microbial degradation of xenobiotic chemical
compounds. Chemical compounds are described in SMILES format and based on a set of
rules that can be found on the UM-PPS webpage, potential pathways for the degradation
of the compound may be predicted. A limitation of this method is the dependence on
biotransformation rules based on known reactions that can be found in the database or
supported by literature.
38
DESHARKY is another pathway prediction tool that generates a candidate pathway and
presents it in the context of the host organism where these novel pathways are to be
implemented.
43
Additionally, amino acid sequences of the enzymes from the closest
organisms phylogenetically related to the host organism are presented. To achieve this,
DESHARKY uses the Monte Carlo algorithm to predict a potential pathway route connecting
two compounds. This tool, however, reconstructs potential pathways from databases of
known enzymatic reactions, and thus is limited in its capability in the bioconversion of
chemical compounds.
43
The KEGG database also has a pathway prediction tool available on its website, PathPred.
39
PathPred utilizes only the metabolic reactions that can be found in the KEGG database to
predict metabolic pathways between two biochemical compounds. Also, the minimum
number of metabolic steps required for the biotransformation can be specified. Again, as
with the UM-PPS tool and DESHARKY, the KEGG pathway prediction tool is limited to
known metabolic reactions and chemical compounds found in the database.
39
One pathway prediction tool that is not limited to known enzymatic/metabolic reactions to
generate pathways between two biochemical compounds is BNICE, in which the chemical
bonds in the reactants that undergo changes during a chemical reaction are represented as a
bond
150
electron matrix (BEM).
2
The BEM is then added with another matrix that represents
a
reaction operator.
The reaction matrix is a representative of the
generalized enzyme
'
'
'
reactions
described by the third-level classification of the Enzyme Commission (EC)
numbers. The method screens out unlikely candidates and predicts the most favorable
pathway by calculating the thermodynamic energy changes along the pathway. Although
BNICE identifies all possible pathways in a given target molecule from a starting molecule,
the novel pathways found using BNICE are devoid of
'
combinatorial explosion,
because it
'
'
weeds out thermodynamically infeasible reactions.
2
Another pathway prediction tool that is not limited to known enzymatic/metabolic
reactions to generate pathways between two biochemical compounds is the standalone
program Enzyme Matcher (EnzMatcher), a retrosynthesis framework with a prioritization
scoring algorithm developed at the Korea Advanced Institute of Science and Technology
(KAIST).
1
This tool employs a set of rules that are not dependent on known metabolic
reactions. Instead, the tool employs a set of rules to determine all possible biochemical
reactions that may occur for each functional group found in the chemical compound. By
using a rule set based on all possible biochemical reactions rather than known metabolic
reactions, EnzMatcher is capable of predicting not only known metabolic reactions that are
found in databases, but also reactions that are biochemically feasible but not characterized
or identified in biological systems. The results from EnzMatcher analysis are then prioritized
based on binding site covalence, chemical similarity of the compounds, thermodynamic
favorability and pathway distance (
Table 8.1
).
1
Although none of the tools aforementioned
have been complemented by experimental validation for their capability of predicting
feasible pathways, the effort to solve such problems is complemented by the development
of tools for
designing enzymes to catalyze chemical reactions that have not yet
been observed in nature.
custom
'
'
