基于机器学习方法虚拟筛选Syk的抑制剂

脾酪氨酸激酶(Syk)是一种细胞内的酪氨酸激酶细胞质受体,在类风湿性关节炎(RA)的发病过程中起着至关重要的作用。筛选Syk抑制剂对RA的治疗有着重要的意义。采用C4.5决策树与随机森林(RF)两种机器学习方法分别对Syk抑制剂与非抑制剂建立模型,经过对比,RF具有更好的预测精度。采用RF模型对Syk抑制剂进行虚拟筛选,从ZINC数据库筛选得到潜在的Syk抑制剂分子。研究结果表明,机器学习方法对于虚拟筛选和发现潜在的Syk抑制剂十分有效。     

天然产物针对Caspase-1的虚拟筛选研究

为了寻找靶向Caspase-1的活性化合物,本研究建立了四种机器学习分类模型(RF、SVM、ANN和VOTING)。这些模型根据不同的评估指标进行了比较,具有高曲线下面积(AUC)的最佳分类模型被用来对ZINC数据库中的天然产物进行虚拟筛选。随后通过不同的药物相似性规则对化合物的ADMET特性进行了过滤。此外,考虑到蛋白质与配体结合的实际情况,本文对每个选定的配体与Caspase-1进行了分子对接和相互作用分析。根据计算出的结合能进行排序,并筛选出3个潜在的抑制剂。     

PPAR激动剂的定向设计、虚拟筛选及合成

过氧化物酶体增殖因子活化受体(PPAR)是核受体超家族的一员.基于受体结构的药物分子设计与组合化学策略相结合,构建了过氧化物酶体增殖因子活化受体(PPAR)激动剂的虚拟化合物库.将已知小分子配体(GW409544)与PPAR晶体复合物进行剥离,得到受体的活性构象,并利用此活性受体分子与虚拟库中小分子进行对接和虚拟筛选,得到理论上结合较强的化合物,并对这些化合物进行合成,共合成9个新化合物.活性筛选结果显示化合物对PPAR具有一定的亲和力,其中有三个化合物显示出对PPARα,PPARγ的双重激动作用,从而指导新活性化合物的设计和合成.     

基于分子描述符和机器学习方法预测和虚拟筛选MMP-13对MMP-1的选择性抑制剂

基质金属蛋白酶-13 (MMP-13)为预防和治疗骨关节炎(OA)提供了充满希望的靶标. 通过抑制剂来阻断MMP-13的活性将会对治疗OA疾病产生潜在的作用. 然而,宽谱抑制剂同样抑制MMP家族的其它成员,特别是MMP-1,这将会导致肌与骨的综合症. 因此,设计和发现潜在的MMP-13 相对于MMP-1 的高效选择性抑制剂,在对治疗OA新型药物的研发中具有相当重要的现实意义. 本研究通过两种机器学习方法(ML)：支持向量机(SVM)和随机森林(RF)来建立分类模型,用于预测不同结构的MMP-13 对MMP-1 的选择性抑制剂. 所建这些模型的预测效果都已经达到了令人满意的精度. 在这两种ML模型中,RF对于MMP-13选择性抑制剂和非抑制剂的精度分别达到97.58%和100%. 同时,与MMP-13对MMP-1的选择性抑制最相关的分子描述符也基于不同的特征选择方法被两种模型挑选出来. 最后,用预测效果最好的RF模型虚拟筛选了ZINC数据库的“fragment-like”子集,从而得到了一系列潜在的候选药物. 研究表明,机器学习方法,特别是RF方法,对于发现潜在的MMP-13选择性抑制剂十分有效. 同时还得到了一些与MMP-13的选择性抑制相关的分子描述符.     

基于分子描述符和机器学习方法预测和虚拟筛选MMP-13对MMP-1的选择性抑制剂

基质金属蛋白酶-13(MMP-13)为预防和治疗骨关节炎(OA)提供了充满希望的靶标.通过抑制剂来阻断MMP-13的活性将会对治疗OA疾病产生潜在的作用.然而,宽谱抑制剂同样抑制MMP家族的其它成员,特别是MMP-1,这将会导致肌与骨的综合症.因此,设计和发现潜在的MMP-13相对于MMP-1的高效选择性抑制剂,在对治疗OA新型药物的研发中具有相当重要的现实意义.本研究通过两种机器学习方法(ML):支持向量机(SVM)和随机森林(RF)来建立分类模型,用于预测不同结构的MMP-13对MMP-1的选择性抑制剂.所建这些模型的预测效果都已经达到了令人满意的精度.在这两种ML模型中,RF对于MMP-13选择性抑制剂和非抑制剂的精度分别达到97.58%和100%.同时,与MMP-13对MMP-1的选择性抑制最相关的分子描述符也基于不同的特征选择方法被两种模型挑选出来.最后,用预测效果最好的RF模型虚拟筛选了ZINC数据库的"fragment-like"子集,从而得到了一系列潜在的候选药物.研究表明,机器学习方法,特别是RF方法,对于发现潜在的MMP-13选择性抑制剂十分有效.同时还得到了一些与MMP-13的选择性抑制相关的分子描述符.     

基于药效团模型的DHODH抑制剂构效关系研究

利用药效团模型研究二氢乳清酸脱氢酶(Dihydroorotate dehydrogenase,DHODH)抑制剂的构效关系,为DHODH抑制剂的虚拟筛选提供新的方法.以31个具有DHODH抑制活性的化合物为训练集化合物,半数抑制浓度(IC50)范围为7~63000 nmol/L,利用Catalyst/HypoGen算法构建DHODH抑制剂药效团模型,通过对训练集化合物多个构象进行叠合,提取药效团特征及三维空间限制构建药效团模型.利用基于CatScramble的交叉验证方法及评价模型对已知活性化合物的活性预测能力,确定较优药效团模型.模型包含1个氢键受体、3个疏水中心,表征了受体配体相互作用时可能发生的氢键相互作用、疏水相互作用和π-π相互作用,4个药效特征在三维空间的排列概括了DHODH抑制剂产生活性的结构特点.所得较优模型对训练集化合物及测试集化合物的计算活性值与实验活性值的相关系数分别为0.8405和0.8788.利用药效团模型对来源于微生物的系列化合物进行虚拟筛选,筛选出59个预测活性较好的化合物,可作为进一步药物研发的候选化合物.     

基于他克林的乙酰胆碱酯酶抑制剂的3D-QSAR研究及虚拟筛选

本文通过对58个他克林派生物乙酰胆碱酯酶抑制剂分子进行建模分析,研究其结构与活性的关系,并通过虚拟筛选方法获得一系列潜在AChE抑制剂双位点分子。首先将一系列他克林二联体化合物与AChE晶体结构对接,获得化合物的活性构象,以此进行建模分析,建立结构与活性之间的三维定量构效关系。所得模型CoMFA、CoMSIA、TopomerCoMFA的交叉验证系数分别为0.510、0.702、0.571,非交叉验证系数为0.998、0.988、0.794,测试集r_(pred)~2为0.750、0.742、0.766,所得模型具有良好的预测性,由此可以为设计高活性的新分子提供理论基础。然后,使用Topomer search对ZINC数据库中的125909分子进行虚拟筛选,得到891个具有潜在AChE抑制活性的分子。最后,对这891个分子进行分子对接,观察分子与晶体结构的结合情况,筛选得到66个具有高选择性的双位点AChE抑制剂分子。     

基于配体的组蛋白去乙酰化酶抑制剂的分子设计

本文采用Topomer Co MFA方法对39个组蛋白去乙酰化酶抑制剂进行了3D-QSAR研究,得到q~2=0. 877,r~2=0. 987的可靠模型。运用基于R基团搜索的Topomer Search技术对ZINC2015数据库进行了虚拟筛选,筛选出一批具有潜在活性的目标化合物,模型预测结果表明,筛选出的化合物活性比最初合成的化合物大幅度提高,其中筛选出的最高活性化合物S2-7(IC_(50)=0. 0235μmol·L~(-1))活性达到了最初合成的高活性化合物21(IC_(50)=0. 103μmol·L~(-1))的4倍。分子对接技术揭示了化合物结构和靶酶之间的联系,为更新型HDACIs的设计以及结构优化提供了重要信息和理论指导。     

ω-芋螺毒素的定量构效关系与虚拟筛选

ω-芋螺毒素属于海洋生物活性多肽,由24-31个氨基酸残基组成.特异性作用于电压敏感的钙离子通道(VGCCs),能够直接开发成药物或作为先导化合物进行新药开发.本文应用新型氨基酸残基结构描述符cscales和遗传偏最小二乘算法,对ω-芋螺毒素进行定量构效关系(QSAR)研究,并设计、构建了容量为2244个化合物的N-型和P/Q-型VGCC拮抗剂虚拟组合多肽库,然后分别采用QSAR模型预测和相似性搜索方法对组合多肽库进行了虚拟筛选.研究结果表明,建立的N-型和P/Q-型VGCC拮抗剂QSAR模型均具有较好的预测能力,交叉验证相关系数(CV-r2)均大于0.89.主成分分析和聚类分析结果表明,虚拟组合多肽库中化合物具有较好的结构多样性和差异性.通过虚拟筛选,得到了具有高预测活性的6个N-型和19个P/Q-型钙离子通道拮抗剂,为进一步的合成和活性评价奠定了理论基础.同时,本文建立的多肽QSAR预测模型和虚拟筛选策略,为其它多肽类化合物的定量构效关系研究和虚拟筛选提供了参考.     

ω-芋螺毒素的定量构效关系与虚拟筛选

ω-芋螺毒素属于海洋生物活性多肽, 由24-31 个氨基酸残基组成. 特异性作用于电压敏感的钙离子通道(VGCCs), 能够直接开发成药物或作为先导化合物进行新药开发. 本文应用新型氨基酸残基结构描述符cscales和遗传偏最小二乘算法, 对ω-芋螺毒素进行定量构效关系(QSAR)研究, 并设计、构建了容量为2244 个化合物的N-型和P/Q-型VGCC拮抗剂虚拟组合多肽库, 然后分别采用QSAR模型预测和相似性搜索方法对组合多肽库进行了虚拟筛选. 研究结果表明, 建立的N-型和P/Q-型VGCC拮抗剂QSAR模型均具有较好的预测能力, 交叉验证相关系数(CV-r²)均大于0.89. 主成分分析和聚类分析结果表明, 虚拟组合多肽库中化合物具有较好的结构多样性和差异性. 通过虚拟筛选, 得到了具有高预测活性的6 个N-型和19 个P/Q-型钙离子通道拮抗剂, 为进一步的合成和活性评价奠定了理论基础. 同时, 本文建立的多肽QSAR预测模型和虚拟筛选策略, 为其它多肽类化合物的定量构效关系研究和虚拟筛选提供了参考.     

Natural Products Extracted from Fungal Species as New Potential Anti-Cancer Drugs: A Structure-Based Drug Repurposing Approach Targeting HDAC7

Mushrooms can be considered a valuable source of natural bioactive compounds with potential polypharmacological effects due to their proven antimicrobial, antiviral, antitumor, and antioxidant activities. In order to identify new potential anticancer compounds, an in-house chemical database of molecules extracted from both edible and non-edible fungal species was employed in a virtual screening against the isoform 7 of the Histone deacetylase (HDAC). This target is known to be implicated in different cancer processes, and in particular in both breast and ovarian tumors. In this work, we proposed the ibotenic acid as lead compound for the development of novel HDAC7 inhibitors, due to its antiproliferative activity in human breast cancer cells (MCF-7). These promising results represent the starting point for the discovery and the optimization of new HDAC7 inhibitors and highlight the interesting opportunity to apply the “drug repositioning” paradigm also to natural compounds deriving from mushrooms.     

Discovery of Monoacylglycerol Lipase (MAGL) Inhibitors Based on a Pharmacophore-Guided Virtual Screening Study

Monoacylglycerol lipase (MAGL) is an important enzyme of the endocannabinoid system that catalyzes the degradation of the major endocannabinoid 2-arachidonoylglycerol (2-AG). MAGL is associated with pathological conditions such as pain, inflammation and neurodegenerative diseases like Parkinson’s and Alzheimer’s disease. Furthermore, elevated levels of MAGL have been found in aggressive breast, ovarian and melanoma cancer cells. Due to its different potential therapeutic implications, MAGL is considered as a promising target for drug design and the discovery of novel small-molecule MAGL inhibitors is of great interest in the medicinal chemistry field. In this context, we developed a pharmacophore-based virtual screening protocol combined with molecular docking and molecular dynamics simulations, which showed a final hit rate of 50% validating the reliability of the in silico workflow and led to the identification of two promising and structurally different reversible MAGL inhibitors, VS1 and VS2. These ligands represent a valuable starting point for structure-based hit-optimization studies aimed at identifying new potent MAGL inhibitors.     

Identification of Human Dihydroorotate Dehydrogenase Inhibitor by a Pharmacophore-Based Virtual Screening Study

Human dihydroorotate dehydrogenase (hDHODH) is an enzyme belonging to a flavin mononucleotide (FMN)-dependent family involved in de novo pyrimidine biosynthesis, a key biological pathway for highly proliferating cancer cells and pathogens. In fact, hDHODH proved to be a promising therapeutic target for the treatment of acute myelogenous leukemia, multiple myeloma, and viral and bacterial infections; therefore, the identification of novel hDHODH ligands represents a hot topic in medicinal chemistry. In this work, we reported a virtual screening study for the identification of new promising hDHODH inhibitors. A pharmacophore-based approach combined with a consensus docking analysis and molecular dynamics simulations was applied to screen a large database of commercial compounds. The whole virtual screening protocol allowed for the identification of a novel compound that is endowed with promising inhibitory activity against hDHODH and is structurally different from known ligands. These results validated the reliability of the in silico workflow and provided a valuable starting point for hit-to-lead and future lead optimization studies aimed at the development of new potent hDHODH inhibitors.     

Prospective virtual screening for novel p53–MDM2 inhibitors using ultrafast shape recognition

The p53 protein, known as the guardian of genome, is mutated or deleted in approximately 50 % of human tumors. In the rest of the cancers, p53 is expressed in its wild-type form, but its function is inhibited by direct binding with the murine double minute 2 (MDM2) protein. Therefore, inhibition of the p53–MDM2 interaction, leading to the activation of tumor suppressor p53 protein presents a fundamentally novel therapeutic strategy against several types of cancers. The present study utilized ultrafast shape recognition (USR), a virtual screening technique based on ligand–receptor 3D shape complementarity, to screen DrugBank database for novel p53–MDM2 inhibitors. Specifically, using 3D shape of one of the most potent crystal ligands of MDM2, MI-63, as the query molecule, six compounds were identified as potential p53–MDM2 inhibitors. These six USR hits were then subjected to molecular modeling investigations through flexible receptor docking followed by comparative binding energy analysis. These studies suggested a potential role of the USR-selected molecules as p53–MDM2 inhibitors. This was further supported by experimental tests showing that the treatment of human colon tumor cells with the top USR hit, telmisartan, led to a dose-dependent cell growth inhibition in a p53-dependent manner. It is noteworthy that telmisartan has a long history of safe human use as an approved anti-hypertension drug and thus may present an immediate clinical potential as a cancer therapeutic. Furthermore, it could also serve as a structurally-novel lead molecule for the development of more potent, small-molecule p53–MDM2 inhibitors against variety of cancers. Importantly, the present study demonstrates that the adopted USR-based virtual screening protocol is a useful tool for hit identification in the domain of small molecule p53–MDM2 inhibitors.     

In silico prediction and screening of γ‐secretase inhibitors by molecular descriptors and machine learning methods

γ‐Secretase inhibitors have been explored for the prevention and treatment of Alzheimer's disease (AD). Methods for prediction and screening of γ‐secretase inhibitors are highly desired for facilitating the design of novel therapeutic agents against AD, especially when incomplete knowledge about the mechanism and three‐dimensional structure of γ‐secretase. We explored two machine learning methods, support vector machine (SVM) and random forest (RF), to develop models for predicting γ‐secretase inhibitors of diverse structures. Quantitative analysis of the receiver operating characteristic (ROC) curve was performed to further examine and optimize the models. Especially, the Youden index (YI) was initially introduced into the ROC curve of RF so as to obtain an optimal threshold of probability for prediction. The developed models were validated by an external testing set with the prediction accuracies of SVM and RF 96.48 and 98.83% for γ‐secretase inhibitors and 98.18 and 99.27% for noninhibitors, respectively. The different feature selection methods were used to extract the physicochemical features most relevant to γ‐secretase inhibition. To the best of our knowledge, the RF model developed in this work is the first model with a broad applicability domain, based on which the virtual screening of γ‐secretase inhibitors against the ZINC database was performed, resulting in 368 potential hit candidates. © 2009 Wiley Periodicals, Inc. J Comput Chem, 2010     

Discovery of New Inhibitors of eEF2K from Traditional Chinese Medicine Based on In Silico Screening and In Vitro Experimental Validation

Eukaryotic elongation factor 2 kinase (eEF2K) is a highly conserved α kinase and is increasingly considered as an attractive therapeutic target for cancer as well as other diseases. However, so far, no selective and potent inhibitors of eEF2K have been identified. In this study, pharmacophore screening, homology modeling, and molecular docking methods were adopted to screen novel inhibitor hits of eEF2K from the traditional Chinese medicine database (TCMD), and then cytotoxicity assay and western blotting were performed to verify the validity of the screen. Resultantly, after two steps of screening, a total of 1077 chemicals were obtained as inhibitor hits for eEF2K from all 23,034 compounds in TCMD. Then, to verify the validity, the top 10 purchasable chemicals were further analyzed. Afterward, Oleuropein and Rhoifolin, two reported antitumor chemicals, were found to have low cytotoxicity but potent inhibitory effects on eEF2K activity. Finally, molecular dynamics simulation, pharmacokinetic and toxicological analyses were conducted to evaluate the property and potential of Oleuropein and Rhoifolin to be drugs. Together, by integrating in silico screening and in vitro biochemical studies, Oleuropein and Rhoifolin were revealed as novel eEF2K inhibitors, which will shed new lights for eEF2K-targeting drug development and anticancer therapy.     

Discovery of novel IDO1 inhibitors via structure-based virtual screening and biological assays

Indoleamine 2,3-dioxygenase 1 (IDO1) is a heme-containing enzyme that catalyzes the first and rate-limiting step in catabolism of tryptophan via the kynurenine pathway, which plays a pivotal role in the proliferation and differentiation of T cells. IDO1 has been proven to be an attractive target for many diseases, such as breast cancer, lung cancer, colon cancer, prostate cancer, etc. In this study, docking-based virtual screening and bioassays were conducted to identify novel inhibitors of IDO1. The cellular assay demonstrated that 24 compounds exhibited potent inhibitory activity against IDO1 at micromolar level, including 8 compounds with IC₅₀ values below 10 μM and the most potent one (compound 1) with IC₅₀ of 1.18?±?0.04 μM. Further lead optimization based on similarity searching strategy led to the discovery of compound 28 as an excellent inhibitor with IC₅₀ of 0.27?±?0.02 μM. Then, the structure–activity relationship of compounds 1, 2, 8 and 14 analogues is discussed. The interaction modes of two compounds against IDO1 were further explored through a Python Based Metal Center Parameter Builder (MCPB.py) molecular dynamics simulation, binding free energy calculation and electrostatic potential analysis. The novel IDO1 inhibitors of compound 1 and its analogues could be considered as promising scaffold for further development of IDO1 inhibitors.