SAR and QSAR in Environmental Research最新文献

This study illustrates the use of chemical fingerprints with machine learning for blood-brain barrier (BBB) permeability prediction. Employing the Blood Brain Barrier Database (B3DB) dataset for BBB permeability prediction, we extracted nine different fingerprints. Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost) algorithms were used to develop models for permeability prediction. Random Forest recursive Feature Selection (RF-RFS) method was used for extracting informative attributes. An additional database was employed for the validation phase. The results indicate that all nine datasets achieved good performance in training, test and validation stages. We further took MACC Keys fingerprints, one of the best performing models for explainability analysis. For this purpose, we used SHapley Additive exPlanations (SHAP) analysis on this dataset for the identification of key structural features influencing BBB permeability prediction. These features include aliphatic carbons, methyl groups and oxygen-containing groups. This study highlights the effectiveness of different fingerprint descriptors in predicting BBB permeability. SHAP analysis provides value additions to the simulations. These simulations will be of significant help in drug discovery processes, particularly in developing Central Nervous System (CNS) therapeutics.

Hepatocellular carcinoma (HCC) ranks fourth in cancer-related mortality worldwide. This study aims to uncover the genes and pathways involved in HCC through network pharmacology (NP) and to discover potential drugs via machine learning (ML)-based ligand screening. Additionally, toxicity prediction, molecular docking, and molecular dynamics (MD) simulations were conducted. NP study identified key genes related to HCC, particularly the enzymes AKT1 and GSK3β. Pathway analysis revealed that crucial pathways like PI3K-AKT and WNT signalling play pivotal roles in HCC progression. Using ML, potential inhibitors for AKT1 and GSK3β were identified, including CHEMBL2177361 and CHEMBL403354 for AKT1, and CHEMBL3652546 and CHEMBL4641631 for GSK3β. post-MD analyses, including RMSD, 2D-RMSD, RMSD cluster, RMSF, PCA, DCCM, residence time analysis, diffusion coefficient, autoencoder-based dimensionality reduction, FEL and MM/GBSA were performed to understand the protein-ligand interactions. The present study reveals the stable interactions of the inhibitors with AKT1 and GSK3β. The binding free energies of all the four complexes were -39.9, -46.8, -41.6, and -45.9 kcal/mol, respectively. This research provides novel insights into the genes and pathways involved in the progression and pathogenesis of HCC using bioinformatics tools. Furthermore, ML-based virtual screening identified potent inhibitors against the target proteins of HCC, such as AKT1 and GSK3β.

ErbB2 kinase is a key target in approximately 20% of breast cancer cases; however, ErbB2-positive cells may shift their dependence to ErbB4 upon developing resistance to ErbB2 inhibitors. Targeting ErbB4 presents a viable strategy to address this challenge. This study employs a comprehensive approach combining structure-based pharmacophore modelling, molecular docking, and MM-GBSA calculations to identify novel ErbB4 kinase inhibitors. Critical pharmacophoric features were extracted from the crystal structures of ErbB4-lapatinib, followed by virtual screening of the Chembl database to discover potential small molecule candidates. Furthermore, the ADMET profiles of 11 shortlisted candidates were assessed to verify their pharmacokinetic and toxicity properties, identifying Chembl310724, Chembl521284, and Chembl4168686 as promising inhibitors of ErbB4 kinase activity with the binding free energy (ΔG_bind) values of -99.84, -89.42 and -86.06 kcal/mol, respectively. This integrated methodology not only enhances our understanding of ErbB4 inhibition but also sets a foundation for the rational design of targeted therapies addressing breast cancer with ErbB4 dependency.

Identifying new compounds with minimal side effects to enhance patients' quality of life is the ultimate goal of drug discovery. Due to the expensive and time-consuming nature of experimental investigations and the scarcity of data in traditional QSAR studies, deep transfer learning models, such as the BERT model, have recently been suggested. This study evaluated the model's performance in predicting the anti-proliferative activity of five cancer cell lines (HeLa, MCF7, MDA-MB231, PC3, and MDA-MB) using over 3,000 synthesized molecules from PubChem. The results indicated that the model could predict the class of designed small molecules with acceptable accuracy for most cell lines, except for PC3 and MDA-MB. The model's performance was further tested on an in-house dataset of approximately 25 small molecules per cell line, based on IC50 values. The model accurately predicted the biological activity class for HeLa with an accuracy of $0.77\pm 0.4$ and demonstrated acceptable performance for MCF7 and MDA-MB231, with accuracy between 0.56 and 0.66. However, the results were less reliable for PC3 and HepG2. In conclusion, the ChemBERTa fine-tuned model shows potential for predicting outcomes on in-house datasets.

CDK/Cyclins are dysregulated in several human cancers. Recent studies showed inhibition of CDK4/6 was responsible for controlling cell cycle progression and cancer cell growth. In the present study, atom-based and field-based 3D-QSAR, virtual screening, molecular docking and molecular dynamics studies were done for the development of novel pyrrolo[2,3-d]pyrimidine (P2P) derivatives as anticancer agents. The developed models showed good Q² and r² values for atom-based 3D-QSAR, which were equal to 0.7327 and 0.8939, whereas for field-based 3D-QSAR the values were 0.8552 and 0.6255, respectively. Molecular docking study showed good-binding interactions with amino acid residues such as VAL-101, HIE-100, ASP-104, ILE-19, LYS-147 and GLU-99, important for CDK4/6 inhibitory activity by using PDB ID: 5L2S. Pharmacophore hypothesis (HHHRR_1) was used in the screening of ZINC database. The top scored ZINC compound ZINC91325512 showed binding interactions with amino acid residues VAL-101, ILE-19, and LYS-147. Enumeration study revealed that the screened compound R1 showed binding interactions with VAL 101 and GLN 149 residues. Furthermore, the Molecular dynamic study showed compound R1, ZINC91325512 and ZINC04000264 having RMSD values of 1.649, 1.733 and 1.610 Å, respectively. These ZINC and enumerated compounds may be used for the development of novel pyrrolo[2,3-d]pyrimidine derivatives as anticancer agent.

Diabetes is attributed to an increased vulnerability to bacterial infection linked to unregulated hyperglycaemia. The present study highlights the formulation of nanoparticles with phyto-compound piperine (PIP) encapsulated within non-toxic biodegradable polymer poly-lactide co-glycolide (PLGA) which showed a variety in surface functionality, biocompatibility, and the ability to tailor an optimized release rate from its polymeric enclosure. The observations revealed that nanopiperine (NPIP) pre-treatment in mice inhibited alteration in hepatic tissue architecture and hepato-biochemical parameters in diabetes and its associated bacterial infections. NPIP also decreased the propensity of lipids to undergo an oxidation process and stabilized the membrane lipids in vivo, thereby lowering oxidative stress and preventing enzymatic activation of CYP1A1. This result is corroborated with the in silico molecular docking study where PIP binding with CYP1A1 gave -11.32 Kcal/mol dock score value. The antibacterial activity of PIP was further demonstrated by the in silico PIP and Ef-Tu protein-binding efficacy revealing -6.48 Kcal/mol score value which was coupled with the results of in vitro studies where the zone of inhibition assay with NPIP against Staphylococcus aureus and Escherichia coli. Thus, NPIP could serve as a potential drug candidate in modulating targeted proteins to inhibit the progression of hyperglycaemia and its associated microbes.

Human aldehyde dehydrogenases (ALDHs) are a group of 19 isoforms often overexpressed in cancer stem cells (CSCs). These enzymes play critical roles in CSC protection, maintenance, cancer progression, therapeutic resistance, and poor prognosis. Thus, targeting ALDH isoforms offers potential for innovative cancer treatments. Flavonoids, known for their ability to affect multiple cancer-related pathways, have shown anticancer activity by downregulating specific ALDH isoforms. This study aimed to evaluate 830 flavonoids from the PubChem database against five ALDH isoforms (ALDH1A1, ALDH1A2, ALDH1A3, ALDH2, ALDH3A1) using computational methods to identify potent inhibitors. Extra precision (XP) Glide docking and MM-GBSA free binding energy calculations identified several flavonoids with high binding affinities. MD simulation highlighted flavonoids 1, 2, 18, 27, and 42 as potential specific inhibitors for each isoform, respectively. Flavonoid 10 showed high binding affinities for ALDH1A2, ALDH1A3, and ALDH3A1, emerging as a potential multi-ALDH inhibitor. ADMET property evaluation indicated that the promising hits have acceptable drug-like profiles, but further optimization is needed to enhance their therapeutic efficacy and reduce toxicity, making them more effective ALDH inhibitors for future cancer treatment.

BACE1 has been regarded as an essential drug design target for treating Alzheimer's disease (AD). Multiple independent Gaussian accelerated molecular dynamics simulations (GaMD), deep learning (DL), and molecular mechanics general Born surface area (MM-GBSA) method are integrated to elucidate the molecular mechanism underlying the effect of D93 and D289 protonation on binding of inhibitors OV6 and 4B2 to BACE1. The GaMD trajectory-based DL successfully identifies significant function domains. Dynamic analysis shows that the protonation of D93 and D289 strongly affects the structural flexibility and dynamic behaviour of BACE1. Free energy landscapes indicate that inhibitor-bound BACE1s have more conformational states in the protonated states than the wild-type (WT) BACE1, and show more binding poses of inhibitors. Binding affinities calculated using the MM-GBSA method indicate that the protonation of D93 and D289 highly disturbs the binding ability of inhibitors to BACE1. In addition, the protonation of two residues significantly affects the hydrogen bonding interactions (HBIs) of OV6 and 4B2 with BACE1, altering their binding activity to BACE1. The binding hot spots of BACE1 recognized by residue-based free energy estimations provide rational targeting sites for drug design towards BACE1. This study is anticipated to provide theoretical aids for drug development towards treatment of AD.

Human polo-like kinase 1 (PLK1) has been recognized as an attractive therapeutic target against nasopharyngeal carcinoma (NPC). The kinase contains a conserved polo-box domain (PBD) that exhibits a wide specificity across various substrates. Previously, we explored natural amino acid preference in PLK1 PBD-binding phosphopeptides. However, limited to the short sequence only natural amino acids cannot guarantee the sufficient exploitation of chemical and structural diversity of the phosphopeptides. Here, we described a genetic optimization (GO) strategy to systematically optimize a 10⁴-sized 6-mer phosphopeptide array towards increasing affinity to PLK1 PBD domain by using 20 natural plus 34 unnatural amino acids as basic building blocks. A QSAR predictor was created to guide the GO optimization and then evaluated rigorously at molecular and cellular levels. Three unnatural phosphopeptides uPP8, uPP15 and uPP20 were designed as potent binders with K_d = 0.18, 0.42 and 0.08 μM, respectively, in which the uPP20 also possessed a good anti-tumor activity against human NPC cells when fused with cell permeation sequence. In addition, we defined a relaxed 6-mer motif for the preferential PLK1 PBD-binding phosphosites, namely [Φ/П]-3-[ζ]-2-[ζ]-1-[pT/pS]0-[Φ/П]+1-[Φ]+2, where the symbols Φ, ζ and П represent hydrophobic, polar and aromatic amino acid types, respectively.  .

Dithiocarbamate fungicides have been widely used in agricultural practices due to their effective control of fungal diseases, thereby contributing to global food security and agricultural productivity. In this study, the inhibitory potency of eight compounds on human and rat aromatase (CYP19A1) activity was evaluated. The results revealed that zineb exhibited the highest inhibitory potency on human CYP19A1 (IC₅₀, 2.79 μM). Maneb (IC₅₀, 3.09 μM), thiram (IC₅₀, 4.76 μM), and ferbam (IC₅₀, 6.04 μM) also demonstrated potent inhibition on human CYP19A1. For the rat CYP19A1, disulfiram (IC₅₀, 1.90 μM) displayed the strongest inhibition followed by maneb (2.16 μM), zineb (2.54 μM), and thiram (6.99 μM). These dithiocarbamates acted as mixed/non-competitive inhibitors of human and rat CYP19A1. Dithiothreitol (DTT), a reducing agent, partially rescued thiram-mediated inhibition when incubated at the same. Moreover, positive correlations were observed between log P, topological polar surface area, molecular weight, and heavy atoms and IC₅₀ values. 3D-QSAR analysis revealed the hydrogen bond acceptor and donor play critical roles in the binding of dithiocarbamates to human CYP19A1. In silico analysis showed that dithiocarbamates bind to the haem binding site, containing Cys437 residues. In conclusion, some dithiocarbamates potently inhibit human and rat CYP19A1 via interacting with haem-binding Cys437 residues.