Abolites serve precise biological functions, we performed an enrichment analysis making use of pathway maps obtained from the KEGG pathway database (http:www.genome.jpkeggpathway.html). We utilized collective and detailed pathway ontologies for the categories “Metabolism,” “Environmental Data Processing,” and “Organismal Systems,” to which the metabolites have been assigned employing chemical structure fingerprints (see Materials and Methods), and calculated the significance of enrichment and depletion for the set of promiscuous and selective metabolites by applying the Fisher’s exact test (Table 4). Regarding metabolism, promiscuous metabolites have been located enriched in energy, nucleotide, and amino acid metabolism pathways. Among the 14 promiscuous metabolites linked with power pathways were energy currency compounds and redox equivalents ADP, ATP, NADH, NAD+ as well because the central metabolites pyruvate, succinate, as well as the amino acid glycine. Partly overlapping with power metabolism, promiscuous compounds had been also identified associated withFrontiers in Molecular Biosciences | www.frontiersin.orgSeptember 2015 | Volume two | ArticleKorkuc and WaltherCompound-protein interactionsFIGURE 8 | Partial least squares regression (PLSR) using physicochemical properties. PLSR prediction models have been built for drug promiscuity (logarithmic pocket count), drug pocket variability and EC entropy of metabolites. (A) Cross-validated (CV) RMSEP (root mean square error of prediction and adjusted CV) curves as function of the number of elements inside the model, (B) loading plot in the physicochemical properties for the very first two elements, and (C) measured against predicted values such as the amount of components applied in the final prediction model (nComp) and correlation coefficient, r, inside a leave-one-out cross-validation setting. PLS models for the respective additional compound classes resulting in inferior performance relative for the 1 shown here are 4-Methylbiphenyl References presented in Supplementary Figures three, 4.Frontiers in Molecular Biosciences | www.frontiersin.orgSeptember 2015 | Volume two | ArticleKorkuc and WaltherCompound-protein interactionsTABLE four | Metabolite pathway, approach, organismal technique ontology enrichment with respect to compound promiscuity. Promiscuous metabolites PFDR -value METABOLISM Collective 4.96E-02 4.96E-02 7.73E-02 Detailed PFDR -value Collective Detailed six.79E-03 three.14E-02 four.52E-02 PFDR -value ORGANISMAL SYSTEMS Collective 4.41E-05 five.42E-04 Detailed 2.68E-02 7.64E-02 Digestive technique Nervous method Vitamin digestion and absorption Synaptic vesicle cycle three.05E-13 Not assigned 1.67E-11 Not assigned Course of action Signal transduction AMPK signaling pathway HIF-1 signaling pathway System PFDR -value Technique Energy metabolism Nucleotide metabolism Amino acid metabolism 6.69E-02 PFDR -value 1.63E-03 1.94E-05 Polyketide sugar unit biosynthesis Procedure Not assigned Not assigned 6.72E-02 9.06E-02 Carbohydrate metabolism Metabolism of terpenoids and polyketides Pathway name PFDR -value Selective metabolites Pathway nameENVIRONMENTAL Details PROCESSINGEnrichment analysis was performed for “Metabolism,” “Environmental Information Processing,” and “Organismal Systems” categories using each collective and detailed ontology terms obtained from the KEGG pathway database. Displayed will be the enriched pathways for promiscuous and selective metabolites with Benjamini-Hochberg procedure corrected p-values (0.1). Note that the category “Not assigned” was introduced for all metabolites.
