Thanks to the fast improvement of the computing power and the rapid development of the computational chemistry and biology, the computer-aided drug design techniques have been successfully applied in almost every stage of the drug discovery and development pipeline to speed up the process of research and reduce the cost and risk related to preclinical and clinical trials. Owing to the development of machine learning theory and the accumulation of pharmacological data, the artificial intelligence (AI) technology, as a powerful data mining tool, has cut a figure in various fields of the drug design, such as virtual screening, activity scoring, quantitative structure-activity relationship (QSAR) analysis, de novo drug design, and in silico evaluation of absorption, distribution, metabolism, excretion and toxicity (ADME/T) properties. Although it is still challenging to provide a physical explanation of the AI-based models, it indeed has been acting as a great power to help manipulating the drug discovery through the versatile frameworks. Recently, due to the strong generalization ability and powerful feature extraction capability, deep learning methods have been employed in predicting the molecular properties as well as generating the desired molecules, which will further promote the application of AI technologies in the field of drug design.

The kinome-wide virtual profiling of small molecules with high-dimensional structure−activity data is a challenging task in drug discovery. Here, we present a virtual profiling model against a panel of 391 kinases based on largescale bioactivity data and the multitask deep neural network algorithm. The obtained model yields excellent internal prediction capability with an auROC of 0.90 and consistently outperforms conventional single-task models on external tests, especially for kinases with insufficient activity data. Moreover, more rigorous experimental validations including 1410 kinasecompound pairs showed a high-quality average auROC of 0.75 and confirmed many novel predicted "off-target" activities. Given the verified generalizability, the model was further applied to various scenarios for depicting the kinome-wide selectivity and the association with certain diseases. Overall, the computational model enables us to create a comprehensive kinome interaction network for designing novel chemical modulators or drug repositioning and is of practical value for exploring previously less studied kinases.

Myeloid-derived suppressor cells (MDSCs) are a prominent component of the pro-tumoral response. The phenotype of and mechanisms used by MDSCs is heterogeneous and requires more precise characterization in gastric cancer (GC) patients. Here, we have identified a novel subset of CD45+CD33lowCD11bdim MDSCs in the peripheral blood of GC patients compared to healthy individuals. CD45+CD33lowCD11bdim MDSCs morphologically resembled neutrophils and expressed high levels of the neutrophil marker CD66b. Circulating CD45+CD33lowCD11bdim MDSCs effectively suppressed CD8+ T cells activity through the inhibition of CD8+ T cell proliferation and interferon-γ (IFN-γ) and granzyme B (GrB) production. The proportion of CD45+CD33lowCD11bdim MDSCs also negatively correlated with the proportion of IFN-γ+CD8+ T cell in the peripheral blood of GC patients. GC patient serum-derived IL-6 and IL-8 activated and induced CD45+CD33lowCD11bdim MDSCs to express arginase I via the PI3K-AKT signaling pathway. This pathway contributed to CD8+ T cell suppression as it was partially rescued by the blockade of the IL-6/IL-8-arginase I axis. Peripheral blood CD45+CD33lowCD11bdim MDSCs, as well as IL-6, IL-8, and arginase I serum levels, positively correlated with GC progression and negatively correlated with overall patient survival. Altogether, our results highlight that a subset of neutrophilic CD45+CD33lowCD11bdim MDSCs is functionally immunosuppressive and activated via the IL-6/IL-8-arginase I axis in GC patients.

The interaction between gastric epithelium and immune response plays key roles in H. pylori–associated pathology. We demonstrated a procolonization and proinflammation role of MMP-10 in H. pylori infection. MMP-10 is elevated in gastric mucosa and is produced by gastric epithelial cells synergistically induced by H. pylori and IL-22 via the ERK pathway. Human gastric MMP-10 was correlated with H. pylori colonization and the severity of gastritis, and mouse MMP-10 from non–BM-derived cells promoted bacteria colonization and inflammation. H. pylori colonization and inflammation were attenuated in IL-22−/−, MMP-10−/−, and IL-22−/−MMP-10−/− mice. MMP-10–associated inflammation is characterized by the influx of CD8+ T cells, whose migration is induced via MMP-10–CXCL16 axis by gastric epithelial cells. Under the influence of MMP-10, Reg3a, E-cadherin, and zonula occludens–1 proteins decrease, resulting in impaired host defense and increased H. pylori colonization. Our results suggest that MMP-10 facilitates H. pylori persistence and promotes gastritis.

Several recent methods have been proposed to obtain significant speed-ups in MRI image reconstruction by leveraging the computational power of GPUs. Previously, we implemented a GPU-based image reconstruction technique called the Illinois Massively Parallel Acquisition Toolkit for Image reconstruction with ENhanced Throughput in MRI (IMPATIENT MRI) for reconstructing data collected along arbitrary 3D trajectories. In this paper, we improve IMPATIENT by removing computational bottlenecks by using a gridding approach to accelerate the computation of various data structures needed by the previous routine. Further, we enhance the routine with capabilities for off-resonance correction and multi-sensor parallel imaging reconstruction. Through implementation of optimized gridding into our iterative reconstruction scheme, speed-ups of more than a factor of 200 are provided in the improved GPU implementation compared to the previous accelerated GPU code.

No abstract

Regulatory T cells (Tregs) are major components of tumor-infiltrating immune cells with potent immunosuppressive properties in gastric cancer (GC) microenvironment. However, different subsets of the Tregs and their relevance to GC are unknown. Here, we found that patients with GC showed a significantly higher Tregs infiltration in tumors, and CD45RA−CCR7− Treg subset constituted most tumor-infiltrating Tregs. Tumor-infiltrating CD45RA−CCR7− Treg subset with an effector/memory phenotype accumulated in tumors and expressed low level of HLA-DR. Gastric tumor-derived TNF-α induced CD45RA−CCR7− Treg subset with similar phenotype to their status in tumors and inhibited their HLA-DR expression via activating STAT3 phosphorylation. These tumor-associated CD45RA−CCR7− Treg subset exerted superior immunosuppressive properties to effectively suppress CD8+ T cells’ anti-tumor function including CD8+ T-cell IFN-γ and granzyme B (GrB) production as well as CD8+ T-cell proliferation in vitro, and also contributed to the growth and progression of human gastric tumors in vivo, via IL-10 secretion and cell–cell contact mechanisms. Moreover, increased tumor-infiltrating CD45RA−CCR7− Treg subset as well as higher intratumoral CD45RA−CCR7− Treg/CD8+ T-cell ratio was associated with advanced disease progression and reduced GC patient survival. This study therefore identifies a novel immunosuppressive pathway involving CD45RA−CCR7− Treg subset development within the GC microenvironment. Efforts to inhibit this pathway may therefore prove a valuable strategy to prevent, and to treat this immune suppressive of GC.

Mast cells are prominent components of solid tumors and exhibit distinct phenotypes in different tumor microenvironments. However, their precise mechanism of communication in gastric cancer remains largely unclear. Here, we found that patients with GC showed a significantly higher mast cell infiltration in tumors. Mast cell levels increased with tumor progression and independently predicted reduced overall survival. Tumor-derived adrenomedullin (ADM) induced mast cell degranulation via PI3K-AKT signaling pathway, which effectively promoted the proliferation and inhibited the apoptosis of GC cells in vitro and contributed to the growth and progression of GC tumors in vivo, and the effect could be reversed by blocking interleukin (IL)-17A production from these mast cells. Our results illuminate a novel protumorigenic role and associated mechanism of mast cells in GC, and also provide functional evidence for these mast cells to prevent, and to treat this immunopathogenesis feature of GC.