Chen Ya, de Bruyn Kops Christina, Kirchmair Johannes
Center for Bioinformatics, Department of Computer Science, Faculty of Mathematics, Informatics and Natural Sciences, Universität Hamburg , Hamburg 20146, Germany.
J Chem Inf Model. 2017 Sep 25;57(9):2099-2111. doi: 10.1021/acs.jcim.7b00341. Epub 2017 Aug 30.
Natural products from plants, animals, marine life, fungi, bacteria, and other organisms are an important resource for modern drug discovery. Their biological relevance and structural diversity make natural products good starting points for drug design. Natural product-based drug discovery can benefit greatly from computational approaches, which are a valuable precursor or supplementary method to in vitro testing. We present an overview of 25 virtual and 31 physical natural product libraries that are useful for applications in cheminformatics, in particular virtual screening. The overview includes detailed information about each library, the extent of its structural information, and the overlap between different sources of natural products. In terms of chemical structures, there is a large overlap between freely available and commercial virtual natural product libraries. Of particular interest for drug discovery is that at least ten percent of known natural products are readily purchasable and many more natural products and derivatives are available through on-demand sourcing, extraction and synthesis services. Many of the readily purchasable natural products are of small size and hence of relevance to fragment-based drug discovery. There are also an increasing number of macrocyclic natural products and derivatives becoming available for screening.
来自植物、动物、海洋生物、真菌、细菌及其他生物体的天然产物是现代药物发现的重要资源。它们的生物学相关性和结构多样性使天然产物成为药物设计的良好起点。基于天然产物的药物发现可从计算方法中大大受益，计算方法是体外测试的一种有价值的前期或补充方法。我们概述了25个虚拟和31个实体天然产物库，这些库在化学信息学应用中，特别是虚拟筛选中很有用。该概述包括每个库的详细信息、其结构信息的范围以及不同天然产物来源之间的重叠情况。就化学结构而言，免费提供的虚拟天然产物库和商业虚拟天然产物库之间有很大的重叠。药物发现特别感兴趣的是，至少10%的已知天然产物易于购买，通过按需采购、提取和合成服务可获得更多的天然产物及其衍生物。许多易于购买的天然产物体积较小，因此与基于片段的药物发现相关。也有越来越多的大环天然产物及其衍生物可用于筛选。
