A deep learning genome-mining strategy for biosynthetic gene cluster prediction

By G. Hannigan, David Prihoda, A. Palicka, Jindrich Soukup, O. Klempíř, Lena Rampula, Jindřich Durčák, M. Wurst, J. Kotowski, Dan Chang and 6 more. Nucleic Acids Research, 2019

325 citations recorded by the source corpus.

Abstract Natural products represent a rich reservoir of small molecule drug candidates utilized as antimicrobial drugs, anticancer therapies, and immunomodulatory agents. These molecules are microbial secondary metabolites synthesized by co-localized genes termed Biosynthetic Gene Clusters (BGCs). The increase in full microbial genomes and similar resources has led to development of BGC prediction algorithms, although their precision and ability to identify novel BGC classes could be improved. Here we present a deep learning strategy (DeepBGC) that offers reduced false positive rates in BGC identification and an improved ability to extrapolate and identify novel BGC classes compared to existing machine-learning tools. We supplemented this with random forest classifiers that accurately predicted BGC product classes and potential chemical activity. Application of DeepBGC to bacterial genomes uncovered previously undetectable putative BGCs that may code for natural products with novel biologic activities. The improved accuracy and classification ability of DeepBGC represents a major addition to in-silico BGC identification.

Subfields: Molecular Biology, Pharmacology

View Paper: https://www.semanticscholar.org/paper/e69e4c4b20da909b048955907cd7a3524bc07999

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Pharmacology

Citations in the corpus

Artificial Intelligence For Natural Product Drug Discovery (2023, methodology): “For example, the hidden Markov model-based method ClusterFinder 15 , the deep learning approaches DeepBGC 16 , GECCO 17 and SanntiS 18 , and several genome mining algorithms for ribosomally synthesized and post-translationally modified peptides (RiPPs) 19–22 each use deep learning”
antiSMASH 6.0: improving cluster detection and comparison capabilities (2021, methodology): “While tools like ARTS (16) and BiG-SCAPE (19) sit downstream of antiSMASH and consume anti-SMASH results, other tools like ClusterFinder (33) and DeepBGC (34) offer alternative, machine-learning based methods to predict gene clusters and thus function more in parallel to antiSMASH’s rule based clust”
MIBiG 3.0: a community-driven effort to annotate experimentally validated biosynthetic gene clusters (2022, methodology): “To address the first stage, sequence information, various tools have been developed that automatically detect BGCs from DNA sequence, including antiSMASH and its siblings fungiSMASH and plantiSMASH (1,2), GECCO (3), DeepBGC (4), RiPPMiner (5) and PRISM 4 (6).”
antiSMASH 7.0: new and improved predictions for detection, regulation, chemical structures and visualisation (2023, unknown): “In turn, anti-SMASH can also incorporate and display BGC predictions from other tools such as DeepBGC ( 19 ) by using the side-loading mechanism introduced in antiSMASH 6 ( 13 ). an-tiSMASH BGC pr edictions ar e included in many genomic and BGC-oriented databases, like the Joint Genome Institute’s…”