A transferable deep learning approach to fast screen potential antiviral drugs against SARS-CoV-2.

The COVID-19 pandemic calls for rapid development of effective treatments. Although various drug repurpose approaches have been used to screen the FDA-approved drugs and drug candidates in clinical phases against SARS-CoV-2, the coronavirus that causes this disease, no magic bullets have been found until now. In this study, we used directed message passing neural network to first build a broad-spectrum anti-beta-coronavirus compound prediction model, which gave satisfactory predictions on newly reported active compounds against SARS-CoV-2. Then, we applied transfer learning to fine-tune the model with the recently reported anti-SARS-CoV-2 compounds and derived a SARS-CoV-2 specific prediction model COVIDVS-3. We used COVIDVS-3 to screen a large compound library with 4.9 million drug-like molecules from ZINC15 database and recommended a list of potential anti-SARS-CoV-2 compounds for further experimental testing. As a proof-of-concept, we experimentally tested seven high-scored compounds that also demonstrated good binding strength in docking studies against the 3C-like protease of SARS-CoV-2 and found one novel compound that can inhibit the enzyme. Our model is highly efficient and can be used to screen large compound databases with millions or more compounds to accelerate the drug discovery process for the treatment of COVID-19.

Factors influencing the healthcare transition in Chinese adolescents with inflammatory bowel disease: a multi-perspective qualitative study.

BACKGROUND: The development and implementation of the transition from pediatric to adult healthcare systems for adolescents with inflammatory bowel disease (IBD) should consider stakeholders' perceptions. This study aimed to explore the factors influencing the transition of Chinese adolescents with IBD from the perspectives of patients, parents, and healthcare providers. METHODS: A descriptive qualitative research was conducted. Purposive sampling was used to recruit 36 participants, including 13 patients, 13 parents, and 10 providers, from a tertiary pediatric IBD center, a tertiary adult IBD center, and the China Crohn's & Colitis Foundation in Zhejiang Province, China. Individual semi-structured interviews were used to collect data on facilitators and barriers to the transition process. Conventional content analysis was used to analyze the interview transcripts. RESULTS: Nine primary themes were identified. Patients with young age, prolonged disease duration, severe disease, academic pressures such as the Gaokao, low level of disease acceptance, limited transition consciousness, low self-efficacy, poor transition communication, and inadequate medical transition system serve as barriers. While patients with the mentality of guilt towards their parents; parents with low education levels and intensive work schedules, high levels of disease acceptance, and situations of parent-child separation; stakeholders with high transition consciousness, high transition self-efficacy, and effective transition communication act as facilitators. Furthermore, community support and hospital guide services were also contributing factors during the transition. CONCLUSIONS: This study offers comprehensive insights into the factors affecting the transition of Chinese adolescent IBD patients. The process is continuously influenced by stakeholders, community, and healthcare environments and policies. Identifying these factors provides healthcare providers with a reference for developing and implementing targeted transition interventions.

Correlation between neutrophil-to-lymphocyte ratio and contrast-induced acute kidney injury and the establishment of machine-learning-based predictive models.

OBJECTIVE: To explore the correlation between neutrophil-to-lymphocyte ratio (NLR) and contrast-induced acute kidney injury (CI-AKI). To develop machine-learning (ML) methods based on NLR and other relevant high-risk factors to establish new and effective predictive models of CI-AKI. Methods: The data of 2230 patients, who underwent elective vascular intervention, coronary angiography and percutaneous coronary intervention were retrospectively collected. The patients were divided into a CI-AKI group and a non-CI-AKI group. Logistic regression was used to analyze the correlation of NLR with CI-AKI and high-risk factors for CI-AKI, and logistic regression (LR), random forest (RF), gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and naïve Bayes (NB) models based on NLR and the high-risk factors were established. RESULTS: A high NLR(>2.844) was an independent risk factor for CI-AKI (odds ratio = 2.304, p < 0.001). The area under the ROC curve (AUC) of the NB model was the largest (0.774), indicating that it had the best performance. NLR, serum creatinine concentration, fasting plasma glucose concentration, and use of ß-blocker all accounted for a large proportion of the predictive performance of each model and were the four most important factors affecting the occurrence of CI-AKI. CONCLUSIONS: There was a significant correlation between NLR and CI-AKI The NB model exhibited the best predictive performance out of the five ML models based on NLR exhibited the best predictive performance out of the five ML models.

A Mechanism Study on the (+)-ESI-TOF/HRMS Fragmentation of Some PPI Prazoles and Their Related Substances.

Fragmentation mechanisms of some prazoles and their related substances were newly investigated in this paper via positive mode ESI-TOF HRMS1 and HRMS2. Some novel fragmentation rules or ions were found or detected in the research. The pyridine and the benzoimidazole ring remained in most cases during the ionization, and heterolytic fragmentations often occurred near the -S(O)nCH2- linker to give the [1,3]-H migration ion or [1,7]-H migration ion rearranging across the benzoimdazole ring. Smiles rearrangement ionizations also frequently occurred, initiated by the attack of the lone pair electrons from the pyridine ring, and the sulfones gave special N-(2-benzoimdazolyl) pyridine ions (11b and 12c) by a direct extraction from SO2, and the thioethers gave similar framework ions (8c, 9c and 10c) via the rearrangement and a further homolytic cleavage of SH radicals. However, the sulfoxides were seldom detected in the corresponding Smiles rearrangement ions during our measurement, and the N'-oxides of the pyridines did not undergo the Smiles rearrangement ionization due to the absence of the lone pair electrons. The 5/6-membered chelating ions with Na+ or K+ were frequently detected as the molecular and further fragment ions. Some novel and interesting fragment ions containing bivalent (8b and 9b), tetravalent (4b, 5c and 6c) or hexavalent (15b and 16b) sulfurs were first reported here.

Prediction of liquid-liquid phase separating proteins using machine learning.

BACKGROUND: The liquid-liquid phase separation (LLPS) of biomolecules in cell underpins the formation of membraneless organelles, which are the condensates of protein, nucleic acid, or both, and play critical roles in cellular function. Dysregulation of LLPS is implicated in a number of diseases. Although the LLPS of biomolecules has been investigated intensively in recent years, the knowledge of the prevalence and distribution of phase separation proteins (PSPs) is still lag behind. Development of computational methods to predict PSPs is therefore of great importance for comprehensive understanding of the biological function of LLPS. RESULTS: Based on the PSPs collected in LLPSDB, we developed a sequence-based prediction tool for LLPS proteins (PSPredictor), which is an attempt at general purpose of PSP prediction that does not depend on specific protein types. Our method combines the componential and sequential information during the protein embedding stage, and, adopts the machine learning algorithm for final predicting. The proposed method achieves a tenfold cross-validation accuracy of 94.71%, and outperforms previously reported PSPs prediction tools. For further applications, we built a user-friendly PSPredictor web server ( http://www.pkumdl.cn/PSPredictor ), which is accessible for prediction of potential PSPs. CONCLUSIONS: PSPredictor could identifie novel scaffold proteins for stress granules and predict PSPs candidates in the human genome for further study. For further applications, we built a user-friendly PSPredictor web server ( http://www.pkumdl.cn/PSPredictor ), which provides valuable information for potential PSPs recognition.

Uncovering the Dominant Motion Modes of Allosteric Regulation Improves Allosteric Site Prediction.

Allostery is an important mechanism that biological systems use to regulate function at a distant site. Allosteric drugs have attracted much attention in recent years due to their high specificity and the possibility of overcoming existing drug-resistant mutations. However, the discovery of allosteric drugs remains challenging as allosteric regulation mechanisms are not clearly understood and allosteric sites cannot be accurately predicted. In this study, we analyzed the dominant modes that determine motion correlations between allosteric and orthosteric sites using the Gaussian network model and found that motion correlations between allosteric and orthosteric sites are dominated by either fast or slow vibrational modes. This dependence of modes results from the relative locations of the two sites and local secondary structures. Based on these analyses, we developed CorrSite2.0 to predict allosteric sites by taking the maximum of the Z-scores calculated from using either slow or fast modes. The prediction accuracy of CorrSite2.0 outperformed other commonly used allosteric site prediction methods with prediction accuracy over 90.0%. Our study uncovers the relationship of protein structure, dynamics, and allosteric regulation and demonstrates that using the dominant motion modes greatly improves allosteric site prediction accuracy. CorrSite2.0 has been integrated into the CavityPlus web server, which can be accessed at http://www.pkumdl.cn/cavityplus. CorrSite2.0 provides a powerful and user-friendly tool for allosteric drug and protein design.

MolMiner: You Only Look Once for Chemical Structure Recognition.

Molecular structures are commonly depicted in 2D printed forms in scientific documents such as journal papers and patents. However, these 2D depictions are not machine readable. Due to a backlog of decades and an increasing amount of printed literatures, there is a high demand for translating printed depictions into machine-readable formats, which is known as Optical Chemical Structure Recognition (OCSR). Most OCSR systems developed over the last three decades use a rule-based approach, which vectorizes the depiction based on the interpretation of vectors and nodes as bonds and atoms. Here, we present a practical software called MolMiner, which is primarily built using deep neural networks originally developed for semantic segmentation and object detection to recognize atom and bond elements from documents. These recognized elements can be easily connected as a molecular graph with a distance-based construction algorithm. MolMiner gave state-of-the-art performance on four benchmark data sets and a self-collected external data set from scientific papers. As MolMiner performed similarly well in real-world OCSR tasks with a user-friendly interface, it is a useful and valuable tool for daily applications. The free download links of Mac and Windows versions are available at https://github.com/iipharma/pharmamind-molminer.

Design, synthesis, and biological evaluation of 2-arylamino-4-(piperidin-4-yloxy)pyrimidines as potent EGFRT790M/L858R inhibitors to treat non-small cell lung cancer.

Three types of 2-arylamino-4-(piperidin-4-yloxy)pyrimidines (I-III) were designed and synthesized as covalent EGFR(epidermal growth factor receptor)T790M/L858R inhibitors by replacement of the common reported 4-(3-amino)phenoxyl moiety with 4-(4-hydroxy)piperidine-4-oxyl, and the introduction of fused-thiophene or -pyrrole on the pyrimidine core to strategically achieve conformational restriction. According to our biological evaluation, it was found that compound 9i could potently suppress EGFRT790M/L858R kinase and H1975 cell proliferation, with IC50 values of 4.902 nM and 0.6210 µM, respectively. Further study showed that 9i not only demonstrated highly selective inhibitory effects toward EGFRT790M/L858R over wild-type EGFR (EGFRWT), but it also had low cytotoxicity against normal HBE(human bronchial epithelial) and L-02 cells. Action mechanism studies showed that 9i effectively hindered cell migration and promoted apoptosis by AO(Acridine Orange)/EB(Ethidium Bromide) staining. These data would provide important clues for the screening of novel covalent EGFRT790M/L858R inhibitors for non-small cell lung cancer (NSCLC) treatment.

Design, synthesis, and biological evaluation of cyano-substituted 2,4-diarylaminopyrimidines as potent JAK3 inhibitors for the treatment of B-cell lymphoma.

A series of cyano-substituted 2,4-diarylaminopyrimidines was designed and synthesized as potent non-covalent JAK3 inhibitors. Among the derivatives synthesized, 9o (IC50 = 22.86 nM), 9 k (IC50 = 21.58 nM), and 9j (IC50 = 20.66 nM) demonstrated inhibitory potencies against JAK3 similar to the known JAK3 inhibitor tofacitinib (IC50 = 20.10 nM). Moreover, 9o displayed potent anti-proliferative activities against Raji and Ramos cells, with IC50 values of 0.9255 µM and 1.405 µM, respectively. In addition, 9o demonstrated low toxicity in normal HBE (human bronchial epithelial cells, IC50 > 10 µΜ) and L-02 (human liver cells, IC50 = 3.104 µΜ) cells. Analysis of the mode of action by flow cytometry indicated that 9o effectively arrested Raji cells at the G2/M phase. Taken together, these results suggested that 9o might be a promising candidate for development as a potential treatment for B-cell lymphoma.

Design, synthesis, and biological evaluation of hydroxamic acid-substituted 2,4-diaryl aminopyrimidines as potent EGFRT790M/L858R inhibitors for the treatment of NSCLC.

A series of 2,4-diarylaminopyrimidine derivatives bearing hydrophilic hydroxamic acids were designed and synthesized as potent EGFRT790M/L858R inhibitors. Among the derivatives synthesized, 10c (IC50 = 5.192 nM), 10j (IC50 = 10.35 nM), and 10o (IC50 = 0.3524 nM) exhibited higher potencies against EGFRT790/M/L858R compared to the known EGFR inhibitor AZD-9291 (IC50 = 20.80 nM). Moreover, 10j showed moderate activity against H1975 cells transfected with the EGFRT790M/L858R mutant, with an IC50 of 0.2113 µM over A431 (wild-type EGFR, SI = 47.3). In addition, 10j exhibited low toxicity in normal HBE cells (human bronchial epithelial cells, IC50 > 40 µΜ). Analysis of the mode of action indicated that 10j effectively induced apoptosis in H1975 cells by arresting the cells in the G2/M phase. Compound 10j also demonstrated efficacy in inhibiting tumor growth in a H1975 xenograft mouse model without losing body weight or killing the mice. Taken together, these results suggested that 10j might be a promising candidate for development as a potential treatment for NSCLC harboring the EGFRT790M/L858R mutation.

Low-Temperature Nickel-Catalyzed C-N Cross-Coupling via Kinetic Resolution Enabled by a Bulky and Flexible Chiral N-Heterocyclic Carbene Ligand.

The transition-metal-catalyzed C-N cross-coupling has revolutionized the construction of amines. Despite the innovations of multiple generations of ligands to modulate the reactivity of the metal center, ligands for the low-temperature enantioselective amination of aryl halides remain a coveted target of catalyst engineering. Designs that promote one elementary reaction often create bottlenecks at other steps. We here report an unprecedented low-temperature (as low as -50 °C), enantioselective Ni-catalyzed C-N cross-coupling of aryl chlorides with sterically hindered secondary amines via a kinetic resolution process (s factor up to >300). A bulky yet flexible chiral N-heterocyclic carbene (NHC) ligand is leveraged to drive both oxidative addition and reductive elimination with low barriers and control the enantioselectivity. Computational studies indicate that the rotations of multiple σ-bonds on the C2 -symmetric chiral ligand adapt to the changing needs of catalytic processes. We expect this design would be widely applicable to diverse transition states to achieve other challenging metal-catalyzed asymmetric cross-coupling reactions.

JAK3 inhibitors based on thieno[3,2-d]pyrimidine scaffold: design, synthesis and bioactivity evaluation for the treatment of B-cell lymphoma.

JAK3 is predominantly expressed in hematopoietic cells and has been a promising therapeutic target for the treatment of B-cell lymphoma. In this study, a new class of thieno[3,2-d]pyrimidines harboring acrylamide pharmacophore were synthesized as potent covalent JAK3 inhibitors (IC50 < 10 nM). Among them, 9a and 9 g displayed the strongest inhibitory potency against JAK3 kinase activity, with IC50 values of 1.9 nM and 1.8 nM, respectively. Furthermore, compared with the reference agents, Spebrutinib and Ibrutinib, 9a not only demonstrated enhanced antiproliferative activity against B lymphoma cells, but also showed very weak proliferative inhibition against normal peripheral blood mononuclear cells (PBMCs) at a concentration of 20 µM. Analysis of the mechanism revealed that 9a could induce the obvious apoptosis in B lymphoma cells and prevent JAK3-STAT3 cascade as well as BTK pathway. Taken together, 9a may be served as a potential new JAK3 inhibitor for the treatment of B-cell lymphoma.

CavityPlus: a web server for protein cavity detection with pharmacophore modelling, allosteric site identification and covalent ligand binding ability prediction.

CavityPlus is a web server that offers protein cavity detection and various functional analyses. Using protein three-dimensional structural information as the input, CavityPlus applies CAVITY to detect potential binding sites on the surface of a given protein structure and rank them based on ligandability and druggability scores. These potential binding sites can be further analysed using three submodules, CavPharmer, CorrSite, and CovCys. CavPharmer uses a receptor-based pharmacophore modelling program, Pocket, to automatically extract pharmacophore features within cavities. CorrSite identifies potential allosteric ligand-binding sites based on motion correlation analyses between cavities. CovCys automatically detects druggable cysteine residues, which is especially useful to identify novel binding sites for designing covalent allosteric ligands. Overall, CavityPlus provides an integrated platform for analysing comprehensive properties of protein binding cavities. Such analyses are useful for many aspects of drug design and discovery, including target selection and identification, virtual screening, de novo drug design, and allosteric and covalent-binding drug design. The CavityPlus web server is freely available at http://repharma.pku.edu.cn/cavityplus or http://www.pkumdl.cn/cavityplus.

A Bulky Chiral N-Heterocyclic Carbene Palladium Catalyst Enables Highly Enantioselective Suzuki-Miyaura Cross-Coupling Reactions for the Synthesis of Biaryl Atropisomers.

Axially chiral biaryl scaffolds are essential structural units in chemistry. The asymmetric Pd-catalyzed Suzuki-Miyaura cross-coupling reaction has been widely recognized as one of the most practical methods for constructing atropisomers of biaryls. However, longstanding challenges remain in this field. For example, substrate scope is often narrow and specialized, functional groups and heterocycles can lead to reduced reactivity and selectivity, bulky ortho-substituents are usually needed, and reported methods are generally inapplicable to tetra-ortho-substituted biaryls. We have developed an unprecedented highly enantioselective N-heterocyclic carbene (NHC)-Pd catalyzed Suzuki-Miyaura cross-coupling reaction for the synthesis of atropisomeric biaryls. These reactions enable efficient coupling of aryl halides (Br, Cl) or aryl triflates with various types of aryl boron compounds (B(OH)2, Bpin, Bneo, BF3K), tolerate a remarkably broad scope of functional groups and heterocycles (>41 examples), employ low loading of catalyst (0.2-2 mol %), and proceed under mild conditions. The protocol provided general and efficient access to various atropisomeric biaryls and heterobiaryls in excellent enantioselectivities (up to 99% ee) with no need of using bulky ortho-substituted substrates and was effective for the synthesis of tetra-ortho-substituent biaryls. Moreover, the method was successfully applied to the diastereo- and enantioselective synthesis of atropisomeric ternaphthalenes. Critical to the success of the reaction is the development and application of an extremely bulky C2-symmetric chiral NHC, (R,R,R,R)-DTB-SIPE, as the ligand for palladium. To the best of our knowledge, this is the first highly enantioselective (>90% ee) example of a chiral NHC-metal-catalyzed C(sp2)-C(sp2) cross-coupling reaction.

Self-Learning Perfect Optical Chirality via a Deep Neural Network.

Optical chirality occurs when materials interact differently with light in a specific circular polarization state. Chiroptical phenomena inspire wide interdisciplinary investigations, which require advanced designs to reach strong chirality for practical applications. The development of artificial intelligence provides a new vision for the manipulation of light-matter interaction beyond the theoretical interpretation. Here, we report a self-consistent framework named the Bayesian optimization and convolutional neural network that combines Bayesian optimization and deep convolutional neural network algorithms to calculate and optimize optical properties of metallic nanostructures. Both electric-field distributions at the near field and reflection spectra at the far field are calculated and self-learned to suggest better structure designs and provide possible explanations for the origin of the optimized properties, which enables wide applications for future nanostructure analysis and design.

Selective synthesis of 3-deoxy-5-hydroxy-1-amino-carbasugars as potential α-glucosidase inhibitors.

A convenient synthesis of novel 3-deoxy-5-hydroxy-1-aminocarbasugars was developed here. The benzyl-protected glucose-derived ketone 12 was selectively converted in high yield to enone 13via retro-Michael elimination of BnOH. The double bond of 13 was regio- and stereo-selectively reduced by the induction of C4-α-OBn to the multi-functionalized 15. 15 contained all the functionalities with similar configurations to carbasugars but with 3-H and 5-OH in the ring, and it would be a very interesting building block for organic synthesis or for bioactive compounds. As one application, 15 was further transformed into 1-amino-carbasugars by the reductive amination and final deprotection of benzyls. The targets were subjected to the in vitro inhibitory activity test against sucrase or maltase. The inhibitory activity of 17b, 17h or 17j against sucrase was nearly similar to that of voglibose. In comparison with voglibose, in vivo results similarly showed that 17b, 17h or 17j could lower the post-prandial blood glucose level after sucrose loading in healthy male ICR mice, while miglitol or acarbose was less effective. The molecular modeling study of some targets or voglibose with human sucrase could explain the inhibiting action.

Efficient molecular encoders for virtual screening.

Molecular representations encoding molecular structure information play critical roles in molecular virtual screening (VS). In order to improve VS performance, an abundance of molecular encoders have been developed and tested by various VS challenges. Combinational strategies were also used to improve the performance. Deep learning (DL)-based molecular encoders have attracted much attention for their automatic information extraction ability. In this review, we present an overview of two-dimensional-, three-dimensional-, and DL-based molecular encoders, summarize recent progress of VS using DL technologies, and propose a general framework of DL molecular encoder-based VS. Perspectives on the future directions of molecular representations and applications in the prediction of active compounds are also provided.

Prediction of Human Cytochrome P450 Inhibition Using a Multitask Deep Autoencoder Neural Network.

Adverse side effects of drug-drug interactions induced by human cytochrome P450 (CYP450) inhibition is an important consideration in drug discovery. It is highly desirable to develop computational models that can predict the inhibitive effect of a compound against a specific CYP450 isoform. In this study, we developed a multitask model for concurrent inhibition prediction of five major CYP450 isoforms, namely, 1A2, 2C9, 2C19, 2D6, and 3A4. The model was built by training a multitask autoencoder deep neural network (DNN) on a large dataset containing more than 13 000 compounds, extracted from the PubChem BioAssay Database. We demonstrate that the multitask model gave better prediction results than that of single-task models, previous reported classifiers, and traditional machine learning methods on an average of five prediction tasks. Our multitask DNN model gave average prediction accuracies of 86.4% for the 10-fold cross-validation and 88.7% for the external test datasets. In addition, we built linear regression models to quantify how the other tasks contributed to the prediction difference of a given task between single-task and multitask models, and we explained under what conditions the multitask model will outperform the single-task model, which suggested how to use multitask DNN models more effectively. We applied sensitivity analysis to extract useful knowledge about CYP450 inhibition, which may shed light on the structural features of these isoforms and give hints about how to avoid side effects during drug development. Our models are freely available at http://repharma.pku.edu.cn/deepcyp/home.php or http://www.pkumdl.cn/deepcyp/home.php .

Prostate-Specific Membrane Antigen Targeted Therapy of Prostate Cancer Using a DUPA-Paclitaxel Conjugate.

Prostate cancer (PCa) is the most prevalent cancer among men in the United States and remains the second-leading cause of cancer mortality in men. Paclitaxel (PTX) is the first line chemotherapy for PCa treatment, but its therapeutic efficacy is greatly restricted by the nonspecific distribution in vivo. Prostate-specific membrane antigen (PSMA) is overexpressed on the surface of most PCa cells, and its expression level increases with cancer aggressiveness, while being present at low levels in normal cells. The high expression level of PSMA in PCa cells offers an opportunity for target delivery of nonspecific cytotoxic drugs to PCa cells, thus improving therapeutic efficacy and reducing toxicity. PSMA has high affinity for DUPA, a glutamate urea ligand. Herein, a novel DUPA-PTX conjugate is developed using DUPA as the targeting ligand to deliver PTX specifically for treatment of PSMA expressing PCa. The targeting ligand DUPA enhances the transport capability and selectivity of PTX to tumor cells via PSMA mediated endocytosis. Besides, DUPA is conjugated with PTX via a disulfide bond, which facilitates the rapid and differential drug release in tumor cells. The DUPA-PTX conjugate exhibits potent cytotoxicity in PSMA expressing cell lines and induces a complete cessation of tumor growth with no obvious toxicity. Our findings give new insight into the PSMA-targeted delivery of chemotherapeutics and provide an opportunity for the development of novel active targeting drug delivery systems for PCa therapy.

The synthesis of 4-arylamido-2-arylaminoprimidines as potent EGFR T790M/L858R inhibitors for NSCLC.

A series of 4-arylamido-2-arylaminoprimidines bearing acrylamide pharmacophore were synthesized as potent EGFRT790M/L858R inhibitors among which 9c (IC50 = 0.5872 nM), 9d (IC50 = 2.213 nM), or 9h (IC50 = 12.57 nM) showed more potent anti-EGFRT790M/L858R activity compared with AZD-9291 (IC50 = 20.80 nM) and possessed high SI displaying 307.6, 56.5, or 12.5 for EGFRT790M/L858R over the wild-type respectively. 9h also showed pretty good activity against H 1975 cells with an IC50 of 1.664 µM and exhibited low toxicity against the normal HBE cells (IC50 > 20 µΜ). 9h had moderate selectivity for H 1975 over A 431 (SI = 7.0) and the other selected cell lines. Morphological staining results further indicated that 9h could promote apoptosis. Hence, 9h was a promising compound for further investigation as a potential EGFRT790M/L858R inhibitor for the treatment of NSCLC.