*2.3. Antibacterial Activity and Parent Ion Distribution*

Culturing all strains in four growth media resulted in 100 metabolite extracts, of which 72% exhibited activity against six pathogenic bacteria (*Escherichia coli*, *Staphylococcus aureus*, *Klebsiella pneumoniae*, *Acinetobacter baumanii*, *Pseudomonas aeruginosa* and *Enterococcus faecalis*) known as the "ESKAPE" pathogens [48]. Specifically, 39% of the biologically active strains (28% of the total number) showed antibacterial activity against only one pathogen. The same percentage of active extracts inhibited the growth of one or more pathogens under only one cultivation condition. The inhibition zones ranged from 0.1 to 2.1 cm (Table S4). Most of the bacterial metabolite extracts were active against *S. aureus* with the *Pseudonocardia* strain KRD185 showing the largest inhibition zone (2.1 cm). Moreover, the 10 fold diluted TSB extract of strain KRD185 showed promising antibacterial activity against *K. pneumoniae* (0.9 cm) and *A. baumannii* (1.5 cm), whereas the A1M1 and 10-fold diluted TSB extracts of *Rhodococcus* strain KRD175 were selectively active against *K. pneumoniae* (Table S4). The occurrence of parent ions in relation to the bioactive extract is shown in Figure 4 where Hinton diagrams are illustrating the number of parent ions produced only by each strain (strain specific), as well as shared between two strains (Figure 4A). Strains belonging to the same genus shared the highest number of produced parent ions, which of course varies between the genera. For example, within the *Micrococcus* genus, strains KRD022 and KRD026 share the highest number of parent ions (649 in total), whereas strains KRD128 and KRD096 share only 287 parent ions. For the *Pseudonocardia* spp. isolates, there is a wide variation in the number of shared parent ions (163–632), where strain KRD176 shares 632 ions with KRD196 and only 163 with KRD291. Variations ranging from 104 to 649 shared parent ions also occurred between strains of different genera, as expected. The *Microbacterium* sp. strain KRD174 shared 104 and 123 parent ions with *Micrococcus* sp. KRD096 and *Pseudonocardia* sp. KRD291, respectively, representing the lowest number of shared ions. On the other hand, the *Rhodococcus* strain KRD175 showed the largest number of shared parent ions (649) with strain KRD022 of the *Micrococcus* genus. The sole *Kocuria* sp. isolate, KRD140, shared only 199 parent ions with *Pseudonocardia* strain KRD176, but shared more than 400 ions with three *Pseudonocardia* strains (KRD182, KRD185, KRD188). Amongst the *Pseudonocardi*a strains, KRD182 showed the highest number of shared parent ions with *Micrococcus* spp. (356–607 parent ions), and *Rhodococcus* strains (465–612 parent ions). The Hinton diagram in Figure 4B demonstrates the number of parent ions produced by strains under each growth condition (white box), as well as the number of specific parent ions per strain grown under each growth condition (purple box) and the observed bioactivity (black outline). The 10-fold diluted TSB metabolite extract of KRD185 (*Pseudonocardia* sp.) had the highest bioactivity, with a zone of inhibition of 2.1 cm, and it produced the highest number of total (803) and unique (29) parent ions in that growth condition when compared to the rest of the studied isolates. Although the *Micrococcus* strain KRD022 produced the highest numbers of parent ions in ISP2 (819) and A1M1 (864) media, it did not exhibit any biological activity against the pathogenic bacteria. These two examples indicate that bioactivity is not necessarily related to the highest number of produced metabolites (parent ions).

#### *2.4. Molecular Networking*

A molecular network of all 100 microbial metabolite extracts (25 strains cultured in four media), in addition to the media and solvent blanks, consisted of 3107 parent ions (nodes). There were 721 nodes that were excluded from the data analysis as they corresponded to parent ions present in the media and solvent blanks. Of the total number of parent ions produced (i.e., not present in the blanks, 2386), 414 of these nodes were singletons indicating that their fragmentation pattern did not correlate with that of any other parent ion, suggesting chemical novelty within the dataset. After ions in the media controls were excluded, 23% (549) of ions were produced by strains grown in all four media. A further 65% (1551) of ions were produced in more than one medium (i.e., not media specific). Interestingly, the percentage of nodes which were media-specific was almost

constant across ISP2 (8.3%), ISP3 (8.8%), and 10-fold diluted TSB (8.0%) media, whereas 6.3% of the produced ions were present in the metabolite extracts derived from A1M1 medium. The MolNetEnhancer workflow showed 38 (putative) chemical classes annotated in the molecular network (Figure 5B and Figure S2). Almost 71% of the produced ions did not match any chemical classification which implies chemical novelty and is in accordance with the low number of library hits generated by the GNPS molecular network. Fatty acyls, and benzene and substituted derivatives represented 4% and 5% of the produced ions, whereas prenol lipids covered 6% of the chemical classes identified in the network.
