Rapid Mapping of All Atoms in Biochemical Reactions


In the design and bioengineering of metabolic pathways for clean bioenergy and other applications, it is important to understand and eventually manipulate the movement of atoms in these biochemical reactions. For example, assessing how a reactant compound is transformed into a targeted product allows researchers to optimize for efficiency in the pathways. A new computational system (Minimum Weighted Edit-Distance or MWED) allows mapping of all the non-hydrogen atoms in biochemical reactions from the initial reactants to the final products. MWED relies on predicting the propensity of forming or breaking chemical bonds during a biochemical reaction. It then calculates and optimizes all possible solutions to the reaction of interest. Because it also uses a mixed-integer linear programming technique, it is three-fold faster than other, similar techniques. The MWED all atom pathway mapping was benchmarked on 2,446 manually curated biochemical reactions from the KEGG database. The researchers found that only 22 MWED-predicted reactions were in error (error rate of 0.9%) due mainly to difficulties in representing stereochemistry in the reactions. MWED offers research scientists an extremely fast and highly accurate method to model all atoms in biochemical reactions, both for novel bioengineering as well as for tracking isotopically labeled atoms in metabolic experiments.


Latendresse, M., J. P. Malerich, M. Travers, and P. D. Karp. 2012. “Accurate Atom-Mapping Computation for Biochemical Reactions,” Journal of Chemical Information and Modeling, DOI: 10.1021/ci3002217.