Computational QSAR model combined molecular descriptors and fingerprints to predict HDAC1 inhibitors

Abstract

> The dynamic balance between acetylation and deacetylation of histones plays a crucial role in the epigenetic regulation of gene expression. It is equilibrated by two families of enzymes: histone acetyltransferases and histone deacetylases (HDACs). HDACs repress transcription by regulating the conformation of the higherorder chromatin structure. HDAC inhibitors have recently become a class of chemical agents for potential treatment of the abnormal chromatin remodeling process involved in certain cancers. In this study, we constructed a large dataset to predict the activity value of HDAC1 inhibitors. Each compound was represented with seven fingerprints, and computational models were subsequently developed to predict HDAC1 inhibitors via five machine learning methods. These methods include naïve Bayes, k-nearest neighbor, C4.5 decision tree, random forest, and support vector machine (SVM) algorithms. The best predicting model was CDK fingerprint with SVM, which exhibited an accuracy of 0.89. This model also performed best in five-fold cross-validation. Some representative substructure alerts responsible for HDAC1 inhibitors were identified by using MoSS in KNIME, which could facilitate the identification of HDAC1 inhibitors. <

Ensemble modeling with machine learning and deep learning to provide interpretable generalized rules for classifying CNS drugs with high prediction power

“…Quantitative structure–activity relationship (QSAR) modeling has been widely used for CNS drug research over the past decade [ 4 , 7 , 8 , 11 , 12 ]. QSAR modeling is used to predict the effectiveness of drug candidates and to provide useful insight to scientists such as structural features for the BBB penetration [ 11 , 13 ].…”

“…A DT is a white-box model that constructs a binary tree of decision nodes that can either use regression or classification. DT’s main advantage lies in the visual interpretability of the trained models, which makes it especially appealing to scientists [ 11 , 12 ]. RF is an ensemble learning algorithm that utilizes the building of multiple DTs.…”

“…RF is an ensemble learning algorithm that utilizes the building of multiple DTs. This process helps circumvent DT’s main drawback of overfitting to the training set, causing RF to be very popular in scientific works [ 11 , 12 ]. With the recent improvement in computing capabilities, a different type of learning called deep learning (DL) has become more popular.…”

Virtual Screening of DPP-4 Inhibitors Using QSAR-Based Artificial Intelligence and Molecular Docking of Hit Compounds to DPP-8 and DPP-9 Enzymes

“…However, this method only provides reliable predictions is limited chemical space. So, a more extensive and more diverse set of data from the ChEMBL database is used to build the global QSAR method [64].…”

Identification of novel leads as potent inhibitors of HDAC3 using ligand-based pharmacophore modeling and MD simulation

“…In the mainstream of drug discovery, the three dimensional quantitative structure relationship (3D-QSAR) and ligand/structure-based pharmacophore modeling approaches have established a reputation towards designing novel, potent and selective inhibitors against HDACs 38 – 42 . For example, the QSAR modeling was performed using k-nearest neighbor (kNN) and support vector machines (SVM) to design the structurally novel bioactive compound against HDAC1 43 , 44 . Pharmacophore-based virtual screening and MD simulations were used to design the novel hydroxamic acids and non-hydroxamate derivatives as potential inhibitors of HDAC2, HDAC4, and HDAC6 45 – 48 .…”