Skip to main content
. Author manuscript; available in PMC: 2018 Jan 17.
Published in final edited form as: Pac Symp Biocomput. 2018;23:56–67.

Fig. 2.

Fig. 2

Pipeline for querying reactions for drug-metabolite pairs. For each drug-metabolite pair, we subtract the drug vector from the metabolite vector to construct a difference vector. We repeat this process for all reactions in the dataset to create a ranked list of reactions most similar to the original drug-metabolite query.