ADCdb: the database of antibody-drug conjugates.

Antibody-drug conjugates (ADCs) are a class of innovative biopharmaceutical drugs, which, via their antibody (mAb) component, deliver and release their potent warhead (a.k.a. payload) at the disease site, thereby simultaneously improving the efficacy of delivered therapy and reducing its off-target toxicity. To design ADCs of promising efficacy, it is crucial to have the critical data of pharma-information and biological activities for each ADC. However, no such database has been constructed yet. In this study, a database named ADCdb focusing on providing ADC information (especially its pharma-information and biological activities) from multiple perspectives was thus developed. Particularly, a total of 6572 ADCs (359 approved by FDA or in clinical trial pipeline, 501 in preclinical test, 819 with in-vivo testing data, 1868 with cell line/target testing data, 3025 without in-vivo/cell line/target testing data) together with their explicit pharma-information was collected and provided. Moreover, a total of 9171 literature-reported activities were discovered, which were identified from diverse clinical trial pipelines, model organisms, patient/cell-derived xenograft models, etc. Due to the significance of ADCs and their relevant data, this new database was expected to attract broad interests from diverse research fields of current biopharmaceutical drug discovery. The ADCdb is now publicly accessible at: https://idrblab.org/adcdb/.

TransFoxMol: predicting molecular property with focused attention.

Predicting the biological properties of molecules is crucial in computer-aided drug development, yet it's often impeded by data scarcity and imbalance in many practical applications. Existing approaches are based on self-supervised learning or 3D data and using an increasing number of parameters to improve performance. These approaches may not take full advantage of established chemical knowledge and could inadvertently introduce noise into the respective model. In this study, we introduce a more elegant transformer-based framework with focused attention for molecular representation (TransFoxMol) to improve the understanding of artificial intelligence (AI) of molecular structure property relationships. TransFoxMol incorporates a multi-scale 2D molecular environment into a graph neural network + Transformer module and uses prior chemical maps to obtain a more focused attention landscape compared to that obtained using existing approaches. Experimental results show that TransFoxMol achieves state-of-the-art performance on MoleculeNet benchmarks and surpasses the performance of baselines that use self-supervised learning or geometry-enhanced strategies on small-scale datasets. Subsequent analyses indicate that TransFoxMol's predictions are highly interpretable and the clever use of chemical knowledge enables AI to perceive molecules in a simple but rational way, enhancing performance.

Identification and assessment of new PIM2 inhibitors for treating hematologic cancers: A combined approach of energy-based virtual screening and machine learning evaluation.

PIM2, part of the PIM kinase family along with PIM1 and PIM3, is often overexpressed in hematologic cancers, fueling tumor growth. Despite its significance, there are no approved drugs targeting it. In response to this challenge, we devised a thorough virtual screening workflow for discovering novel PIM2 inhibitors. Our process includes molecular docking and diverse scoring methods like molecular mechanics generalized born surface area, XGBOOST, and DeepDock to rank potential inhibitors by binding affinities and interaction potential. Ten compounds were selected and subjected to an adequate evaluation of their biological activity. Compound 2 emerged as the most potent inhibitor with an IC50 of approximately 135.7 nM. It also displayed significant activity against various hematological cancers, including acute myeloid leukemia, mantle cell lymphoma, and anaplastic large cell lymphoma (ALCL). Molecular dynamics simulations elucidated the binding mode of compound 2 with PIM2, offering insights for drug development. These results highlight the reliability and efficacy of our virtual screening workflow, promising new drugs for hematologic cancers, notably ALCL.

Discovery of a Highly Potent and Selective BRD9 PROTAC Degrader Based on E3 Binder Investigation for the Treatment of Hematological Tumors.

BRD9 is a pivotal epigenetic factor involved in cancers and inflammatory diseases. Still, the limited selectivity and poor phenotypic activity of targeted agents make it an atypically undruggable target. PROTAC offers an alternative strategy for overcoming the issue. In this study, we explored diverse E3 ligase ligands for the contribution of BRD9 PROTAC degradation. Through molecular docking, binding affinity analysis, and structure-activity relationship study, we identified a highly potent PROTAC E5, with excellent BRD9 degradation (DC50 = 16 pM) and antiproliferation in MV4-11 cells (IC50 = 0.27 nM) and OCI-LY10 cells (IC50 = 1.04 nM). E5 can selectively degrade BRD9 and induce cell cycle arrest and apoptosis. Moreover, the therapeutic efficacy of E5 was confirmed in xenograft tumor models, accompanied by further RNA-seq analysis. Therefore, these results may pave the way and provide the reference for the discovery and investigation of highly effective PROTAC degraders.

Rational Identification of Novel Antibody-Drug Conjugate with High Bystander Killing Effect against Heterogeneous Tumors.

Bystander-killing payloads can significantly overcome the tumor heterogeneity issue and enhance the clinical potential of antibody-drug conjugates (ADC), but the rational design and identification of effective bystander warheads constrain the broader implementation of this strategy. Here, graph attention networks (GAT) are constructed for a rational bystander killing scoring model and ADC construction workflow for the first time. To generate efficient bystander-killing payloads, this model is utilized for score-directed exatecan derivatives design. Among them, Ed9, the most potent payload with satisfactory permeability and bioactivity, is further used to construct ADC. Through linker optimization and conjugation, novel ADCs are constructed that perform excellent anti-tumor efficacy and bystander-killing effect in vivo and in vitro. The optimal conjugate T-VEd9 exhibited therapeutic efficacy superior to DS-8201 against heterogeneous tumors. These results demonstrate that the effective scoring approach can pave the way for the discovery of novel ADC with promising bystander payloads to combat tumor heterogeneity.

A hybrid energy-based and AI-based screening approach for the discovery of novel inhibitors of JAK3.

The JAKs protein family is composed of four isoforms, and JAK3 has been regarded as a druggable target for the development of drugs to treat various diseases, including hematologic tumors, cancer, and neuronal death. Therefore, the discovery of JAK3 inhibitors with novel scaffolds possesses the potential to provide additional options for drug development. This article presents a structure-based hybrid high-throughput virtual screening (HTVS) protocol as well as the DeepDock algorithm, which is based on geometric deep learning. These techniques were used to identify inhibitors of JAK3 with a novel sketch from a specific "In-house" database. Using molecular docking with varying precision, MM/GBSA, geometric deep learning scoring, and manual selection, 10 compounds were obtained for subsequent biological evaluation. One of these 10 compounds, compound 8, was found to have inhibitory potency against JAK3 and the MOLM-16 cell line, providing a valuable lead compound for further development of JAK3 inhibitors. To gain a better understanding of the interaction between compound 8 and JAK3, molecular dynamics (MD) simulations were conducted to provide more details on the binding conformation of compound 8 with JAK3 to guide the subsequent structure optimization. In this article, we achieved compound 8 with a novel sketch possessing inhibitory bioactivity against JAK3, and it would provide an acceptable "hit" for further structure optimization and modification to develop JAK3 inhibitors.

Discovery of 2-Aminopyrimidine Derivatives as Potent Dual FLT3/CHK1 Inhibitors with Significantly Reduced hERG Inhibitory Activities.

FLT3 inhibitors as single agents have limited effects because of acquired and adaptive resistance and the cardiotoxicity related to human ether-a-go-go-related gene (hERG) channel blockade further impedes safe drugs to the market. Inhibitors having potential to overcome resistance and reduce hERG affinity are highly demanded. Here, we reported a dual FLT3/CHK1 inhibitor 18, which displayed potencies to overcome varying acquired resistance in BaF3 cells with FLT3-TKD and FLT3-ITD-TKD mutations. Moreover, 18 displayed high selectivity over c-KIT more than 1700-fold and greatly reduced hERG affinity, with an IC50 value of 58.4 µM. Further mechanistic studies demonstrated 18 can upregulate p53 and abolish the outgrowth of adaptive resistant cells. In the in vivo studies, 18 demonstrated favorable PK profiles and good safety, suppressed the tumor growth in the MV-4-11 cell inoculated mouse xenograft model, and prolonged the survival in the Molm-13 transplantation model, supporting its further development.

Identification of Potential JNK3 Inhibitors: A Combined Approach Using Molecular Docking and Deep Learning-Based Virtual Screening.

JNK3, a member of the MAPK family, plays a pivotal role in mediating cellular responses to stress signals, with its activation implicated in a myriad of inflammatory conditions. While JNK3 holds promise as a therapeutic target for neurodegenerative disorders such as Huntington's, Parkinson's, and Alzheimer's diseases, there remains a gap in the market for effective JNK3 inhibitors. Despite some pan-JNK inhibitors reaching clinical trials, no JNK-targeted therapies have achieved market approval. To bridge this gap, our study introduces a sophisticated virtual screening approach. We begin with an energy-based screening, subsequently integrating a variety of rescoring techniques. These encompass glide docking scores, MM/GBSA, and artificial scoring mechanisms such as DeepDock and advanced Graph Neural Networks. This virtual screening workflow is designed to evaluate and identify potential small-molecule inhibitors with high binding affinity. We have implemented a virtual screening workflow to identify potential candidate molecules. This process has resulted in the selection of ten molecules. Subsequently, these ten molecules have undergone biological activity evaluation to assess their potential efficacy. Impressively, molecule compound 6 surfaced as the most promising, exhibiting a potent kinase inhibitory activity marked by an IC50 of 130.1 nM and a notable reduction in TNF-α release within macrophages. This suggests that compound 6 could potentially serve as an effective inhibitor for the treatment of neuroinflammation and neurodegenerative diseases. The prospect of further medicinal modifications to optimize compound 6 presents a promising avenue for future research and development in this field. Utilizing binding pose metadynamics coupled with molecular dynamics simulations, we delved into the explicit binding mode of compound 6 to JNK3. Such insights pave the way for refined drug development strategies. Collectively, our results underscore the efficacy of the hybrid virtual screening workflow in the identification of robust JNK3 inhibitors, holding promise for innovative treatments against neuroinflammation and neurodegenerative disorders.

Discovery of potential WEE1 inhibitors via hybrid virtual screening.

G2/M cell cycle checkpoint protein WEE1 kinase is a promising target for inhibiting tumor growth. Although various WEE1 inhibitors have entered clinical investigations, their therapeutic efficacy and safety profile remain unsatisfactory. In this study, we employed a comprehensive virtual screening workflow, which included Schrödinger-Glide molecular docking at different precision levels, as well as the utilization of tools such as MM/GBSA and Deepdock to predict the binding affinity between targets and ligands, in order to identify potential WEE1 inhibitors. Out of ten molecules screened, 50% of these molecules exhibited strong inhibitory activity against WEE1. Among them, compounds 4 and 5 showed excellent inhibitory activity with IC50 values of 1.069 and 3.77 nM respectively, which was comparable to AZD1775. Further investigations revealed that compound 4 displayed significant anti-proliferative effects in A549, PC9, and HuH-7 cells and could also induce apoptosis and G1 phase arrest in PC9 cells. Additionally, molecular dynamics simulations unveiled the binding details of compound 4 with WEE1, notably the crucial hydrogen bond interactions formed with Cys379. In summary, this comprehensive virtual screening workflow, combined with in vitro testing and computational modeling, holds significant importance in the development of promising WEE1 inhibitors.