Artificial intelligence for natural product drug discovery.

Developments in computational omics technologies have provided new means to access the hidden diversity of natural products, unearthing new potential for drug discovery. In parallel, artificial intelligence approaches such as machine learning have led to exciting developments in the computational drug design field, facilitating biological activity prediction and de novo drug design for molecular targets of interest. Here, we describe current and future synergies between these developments to effectively identify drug candidates from the plethora of molecules produced by nature. We also discuss how to address key challenges in realizing the potential of these synergies, such as the need for high-quality datasets to train deep learning algorithms and appropriate strategies for algorithm validation.

antiSMASH 7.0: new and improved predictions for detection, regulation, chemical structures and visualisation.

Microorganisms produce small bioactive compounds as part of their secondary or specialised metabolism. Often, such metabolites have antimicrobial, anticancer, antifungal, antiviral or other bio-activities and thus play an important role for applications in medicine and agriculture. In the past decade, genome mining has become a widely-used method to explore, access, and analyse the available biodiversity of these compounds. Since 2011, the 'antibiotics and secondary metabolite analysis shell-antiSMASH' (https://antismash.secondarymetabolites.org/) has supported researchers in their microbial genome mining tasks, both as a free to use web server and as a standalone tool under an OSI-approved open source licence. It is currently the most widely used tool for detecting and characterising biosynthetic gene clusters (BGCs) in archaea, bacteria, and fungi. Here, we present the updated version 7 of antiSMASH. antiSMASH 7 increases the number of supported cluster types from 71 to 81, as well as containing improvements in the areas of chemical structure prediction, enzymatic assembly-line visualisation and gene cluster regulation.

MIBiG 3.0: a community-driven effort to annotate experimentally validated biosynthetic gene clusters.

With an ever-increasing amount of (meta)genomic data being deposited in sequence databases, (meta)genome mining for natural product biosynthetic pathways occupies a critical role in the discovery of novel pharmaceutical drugs, crop protection agents and biomaterials. The genes that encode these pathways are often organised into biosynthetic gene clusters (BGCs). In 2015, we defined the Minimum Information about a Biosynthetic Gene cluster (MIBiG): a standardised data format that describes the minimally required information to uniquely characterise a BGC. We simultaneously constructed an accompanying online database of BGCs, which has since been widely used by the community as a reference dataset for BGCs and was expanded to 2021 entries in 2019 (MIBiG 2.0). Here, we describe MIBiG 3.0, a database update comprising large-scale validation and re-annotation of existing entries and 661 new entries. Particular attention was paid to the annotation of compound structures and biological activities, as well as protein domain selectivities. Together, these new features keep the database up-to-date, and will provide new opportunities for the scientific community to use its freely available data, e.g. for the training of new machine learning models to predict sequence-structure-function relationships for diverse natural products. MIBiG 3.0 is accessible online at https://mibig.secondarymetabolites.org/.

Navigating and expanding the roadmap of natural product genome mining tools.

Natural products are structurally highly diverse and exhibit a wide array of biological activities. As a result, they serve as an important source of new drug leads. Traditionally, natural products have been discovered by bioactivity-guided fractionation. The advent of genome sequencing technology has resulted in the introduction of an alternative approach towards novel natural product scaffolds: Genome mining. Genome mining is an in-silico natural product discovery strategy in which sequenced genomes are analyzed for the potential of the associated organism to produce natural products. Seemingly universal biosynthetic principles have been deciphered for most natural product classes that are used to detect natural product biosynthetic gene clusters using pathway-encoded conserved key enzymes, domains, or motifs as bait. Several generations of highly sophisticated tools have been developed for the biosynthetic rule-based identification of natural product gene clusters. Apart from these hard-coded algorithms, multiple tools that use machine learning-based approaches have been designed to complement the existing genome mining tool set and focus on natural product gene clusters that lack genes with conserved signature sequences. In this perspective, we take a closer look at state-of-the-art genome mining tools that are based on either hard-coded rules or machine learning algorithms, with an emphasis on the confidence of their predictions and potential to identify non-canonical natural product biosynthetic gene clusters. We highlight the genome mining pipelines' current strengths and limitations by contrasting their advantages and disadvantages. Moreover, we introduce two indirect biosynthetic gene cluster identification strategies that complement current workflows. The combination of all genome mining approaches will pave the way towards a more comprehensive understanding of the full biosynthetic repertoire encoded in microbial genome sequences.

Atropopeptides are a Novel Family of Ribosomally Synthesized and Posttranslationally Modified Peptides with a Complex Molecular Shape.

Biomacromolecules are known to feature complex three-dimensional shapes that are essential for their function. Among natural products, ambiguous molecular shapes are a rare phenomenon. The hexapeptide tryptorubin A can adopt one of two unusual atropisomeric configurations. Initially hypothesized to be a non-ribosomal peptide, we show that tryptorubin A is the first characterized member of a new family of ribosomally synthesized and posttranslationally modified peptides (RiPPs) that we named atropopeptides. The sole modifying enzyme encoded in the gene cluster, a cytochrome P450 monooxygenase, is responsible for the atropospecific formation of one carbon-carbon and two carbon-nitrogen bonds. The characterization of two additional atropopeptide biosynthetic pathways revealed a two-step maturation process. Atropopeptides promote pro-angiogenic cell functions as indicated by an increase in endothelial cell proliferation and undirected migration. Our study expands the biochemical space of RiPP-modifying enzymes and paves the way towards the chemoenzymatic utilization of atropopeptide-modifying P450s.

Hidden Treasures: Microbial Natural Product Biosynthesis off the Beaten Path.

Microbes produce structurally diverse natural products to interact with their environment. Many of the biosynthetic products involved in this "metabolic small talk" have been exploited for the treatment of various diseases. As an alternative to the traditional bioactivity-guided workflow, genome mining has been introduced for targeted natural product discovery based on genome sequence information. In this commentary, we will discuss the evolution of genome mining, as well as its current limitations. The Helfrich laboratory aims to play a leading role in overcoming these limitations with the development of computational strategies to identify noncanonical biosynthetic pathways and to decipher the principles that govern the production of the associated metabolites. We will use these insights to develop algorithms for the prediction of natural product scaffolds. These studies will pave the way toward a more comprehensive understanding of the full biosynthetic repertoire encoded in microbial genomes and provide access to novel metabolites.