Background: Polyketide secondary metabolites have been a rich source of antibiotic discovery for decades. Thousands of novel polyketide synthase (PKS) gene clusters have been identified in recent years with advances in DNA sequencing. However, experimental characterization of novel and useful PKS activities remains complicated. As a result, computational tools to analyze sequence data are essential to identify and prioritize potentially novel PKS activities. Here we exploit the concept of genetically-encoded self-resistance to identify and rank biosynthetic gene clusters for their potential to encode novel antibiotics. Results: To identify PKS genes that are likely to produce an antibacterial compound, we developed an automated method to identify and catalog clusters that harbor potential self-resistance genes. We manually curated a list of known self-resistance genes and searched all NCBI genome databases for homologs of these self-resistance genes in biosynthetic gene clusters. The algorithm takes into account (1) the distance of the potential self-resistance gene to a core enzyme in the biosynthetic gene cluster; (2) the presence of a duplicated housekeeping copy of the self-resistance gene; (3) the presence of close homologs of the biosynthetic gene cluster in diverse species also harboring the putative self-resistance gene; (4) evidence for coevolution of the self-resistance gene and core biosynthetic gene; and (5) self-resistance gene ubiquity. We generated a catalog of 190 unique PKS clusters whose products likely target known enzymes of antibacterial importance. We also present an expanded set of putative self-resistance genes that may be useful in identifying small molecules active against novel microbial targets. Conclusions: We developed a bioinformatic approach to identify and rank biosynthetic gene clusters that likely harbor self-resistance genes and may produce compounds with antibacterial properties. We compiled a list of putative self-resistance genes for novel antibacterial targets, and of orphan PKS clusters harboring these targets. These catalogues are a resource for discovery of novel antibiotics.

中文翻译：

---

鉴定具有自我抗性靶基因的聚酮化合物生物合成基因簇

背景：几十年来，聚酮化合物的次生代谢产物一直是发现抗生素的丰富资源。近年来，随着DNA测序的进步，已经发现了成千上万的新型聚酮化合物合酶（PKS）基因簇。但是，新型和有用的PKS活动的实验表征仍然很复杂。结果，分析序列数据的计算工具对于识别潜在的新型PKS活动并确定其优先级至关重要。在这里，我们利用基因编码的自我抗性的概念来识别和排序生物合成基因簇，以为其编码新型抗生素的潜力。结果：为了鉴定可能产生抗菌化合物的PKS基因，我们开发了一种自动方法来鉴定和分类具有潜在自抗性基因的簇。我们手动整理了一系列已知的自抗性基因，并在所有NCBI基因组数据库中搜索了这些自抗性基因在生物合成基因簇中的同源物。该算法考虑到（1）生物合成基因簇中潜在的自抗性基因与核心酶的距离；（2）存在自我抗性基因的重复客房保管副本；（3）在各种物种中存在生物合成基因簇的紧密同源物，也具有推定的自抗性基因；（4）自抗基因和核心生物合成基因共同进化的证据；（5）自抗基因无处不在。我们生成了190个独特的PKS簇的目录，这些簇的产品可能靶向已知的具有重要抗菌作用的酶。我们还提出了一个扩展的推定的自我抗性基因集，可用于识别对新型微生物靶有活性的小分子。结论：我们开发了一种生物信息学方法来鉴定和排序可能具有自我抗性基因并可能产生具有抗菌特性的化合物的生物合成基因簇。我们编辑了新型抗菌靶标的推定自我抗性基因清单，以及包含这些靶标的孤立PKS簇。这些目录是发现新型抗生素的资源。我们编辑了新型抗菌靶标的推定自我抗性基因清单，以及包含这些靶标的孤立PKS簇。这些目录是发现新型抗生素的资源。我们编辑了新型抗菌靶标的推定自我抗性基因清单，以及包含这些靶标的孤立PKS簇。这些目录是发现新型抗生素的资源。