DeepMHADTA: Prediction of Drug-Target Binding Affinity Using Multi-Head Self-Attention and Convolutional Neural Network

Drug-target interactions provide insight into the drug-side effects and drug repositioning. However, wet-lab biochemical experiments are time-consuming and labor-intensive, and are insufficient to meet the pressing demand for drug research and development. With the rapid advancement of deep learning, computational methods are increasingly applied to screen drug-target interactions. Many methods consider this problem as a binary classification task (binding or not), but ignore the quantitative binding affinity. In this paper, we propose a new end-to-end deep learning method called DeepMHADTA, which uses the multi-head self-attention mechanism in a deep residual network to predict drug-target binding affinity. On two benchmark datasets, our method outperformed several current state-of-the-art methods in terms of multiple performance measures, including mean square error (MSE), consistency index (CI), rm2, and PR curve area (AUPR). The results demonstrated that our method achieved better performance in predicting the drug–target binding affinity.     

VEGF/VEGFR-Targeted Therapy and Immunotherapy in Non-small Cell Lung Cancer: Targeting the Tumor Microenvironment

Non-small cell lung cancer (NSCLC) is the leading cause of death by cancer worldwide. Despite developments in therapeutic approaches for the past few decades, the 5-year survival rate of patients with NSCLC remains low. NSCLC tumor is a complex, heterogeneous microenvironment, comprising blood vessels, cancer cells, immune cells, and stroma cells. Vascular endothelial growth factors (VEGFs) are a major mediator to induce tumor microvasculature and are associated with the progression, recurrence, and metastasis of NSCLC. Current treatment medicines targeting VEGF/VEGF receptor (VEGFR) pathway, including neutralizing antibodies to VEGF or VEGFR and receptor tyrosine kinase inhibitors, have shown good treatment efficacy in patients with NSCLC. VEGF is not only an important angiogenic factor but also an immunomodulator of tumor microenvironment (TME). VEGFs can suppress antigen presentation, stimulate activity of regulatory T (Treg) cells, and tumor-associated macrophages, which in turn promote an immune suppressive microenvironment in NSCLC. The present review focuses on the angiogenic and non-angiogenic functions of VEGF in NSCLC, especially the interaction between VEGF and the cellular components of the TME. Additionally, we discuss recent preclinical and clinical studies to explore VEGF/VEGFR-targeted compounds and immunotherapy as novel approaches targeting the TME for the treatment of NSCLC.     

基于GIS和RS的赣江上游流域土地利用动态趋势分析

利用赣江上游流域1990、1995、2000、2005和2010年5期的TM影像以及2010年的SPOT影像资料,通过目视解译方法及GIS软件的空间叠加分析功能,获取了研究区土地利用覆被变化数据。在此基础上,结合1990—2010年赣江上游流域的统计资料利用灰色关联分析和主成分分析法进行分析,找出引起土地利用变化的主导因素。结果表明:研究区20年来林地和建设用地总量增加,耕地、草地、水域和未利用地总量减少。相同时期不同类型的土地利用变化的驱动因子不相同,不同时期同一类型土地利用变化的驱动因子也不相同。再利用GM(1,1)灰色预测模型预测了2015—2030年间研究区驱动因素和土地利用情况,研究表明主导因素对土地利用变化的影响与1990—2010年间变化趋势一致。     

武夷山不同林龄甜槠林土壤呼吸特征及影响因素

为揭示中亚热带常绿阔叶林群落优势种一甜槠天然林不同林龄林下土壤呼吸(Soil respiration,RS)差异及影响因素,采用LI-8100开路式土壤碳通量系统对武夷山自然保护区不同林龄(18、36、54、72 a)天然甜槠林进行了1年的野外原位测定。结果表明:(1)不同林龄甜槠林RS季节动态呈现明显的单峰趋势,林龄对冬季RS影响并不显著(P>0.05),秋季18 a甜槠林RS与其他3种林龄差异显著(P<0.05),林龄对土壤含水率的季节变化没有显著影响(P>0.05);(2)不同林龄甜槠林5 cm深土壤温度与RS拟合R2明显高于土壤含水率与RS拟合R2,随着林龄增大,RS温度敏感性指数Q10值呈上升趋势,依次为1.551、1.589、1.640、1.664,且54、72 a甜槠林RS温度敏感性指数Q10值显著高于18、36 a(P<0.05);(3)土壤含水率与5 cm深土壤温度共同解释了RS变异的86%—90.3%;0—60 cm土层根系生物量与5 cm深土壤温度共同解释了RS变异的88.3%—91.8%,由此可见,生物因子与非生物因子双因素拟合可以更好地解释不同林龄RS差异。在对未来森林植被土壤呼吸及碳汇功能进行研究时,应在考虑林龄及季节差异的基础上,加强对生物因子的测定。     

豫东平原聚落景观格局变化

中国的快速城镇化必然导致聚落景观格局的变化,但目前相关研究对平原地区关注相对较少。文中选取豫东平原地区的开封、商丘、周口市,运用GIS空间分析方法和ENVI遥感影像处理技术,对该地区1972、2015年市域中观尺度上的聚落规模、聚落景观空间变化特征进行分析。结果显示:(1)在聚落发展过程中,1972—2015年聚落斑块数量有所减少,聚落规模逐渐扩张,市域内最大聚落斑块扩张相对更为明显;对聚落斑块扩张强度的分析发现,城市市辖区及周边乡镇聚落扩张强度相对较高。(2)地区内聚落斑块空间分布表现出聚集分布特征,但该时期聚集程度有所减弱;两个年份聚落核密度分布格局大致相似,局部地区存在多核扩散现象,市域尺度上的核密度分布存在地区差异。(3)对斑块形状指数的分析发现,研究区狭长或曲折聚落斑块在空间上分散布局,整体上区域内聚落形状趋于规则。(4)随着距河流、道路距离的增加,聚落斑块总面积和数量有所减少,且道路对聚落分布仅在一定范围内存在较大影响,距中心城市0—6km范围内聚落受中心城区发展辐射影响较大。本文的分析可为平原地区聚落景观的优化布局和聚落用地的集约化发展提供一定的参考。     

Overexpression of Kdr in adult endocardium induces endocardial neovascularization and improves heart function after myocardial infarction

Dear Editor, Coronary artery disease is a leading cause of mortality and morbidity worldwide.Blockade of effective blood flow to heart muscles results in cardi...     

Genome-wide CRISPR screen identifies synthetic lethality between DOCK1 inhibition and metformin in liver cancer

Metformin is currently a strong candidate anti-tumor agent in multiple cancers. However, its anti-tumor effectiveness varies among different cancers or subpopulations, potentially due to tumor heterogeneity. It thus remains unclear which hepatocellular carcinoma (HCC) patient subpopulation(s) can benefit from metformin treatment. Here, through a genome-wide CRISPR-Cas9-based knockout screen, we find that DOCK1 levels determine the anti-tumor effects of metformin and that DOCK1 is a synthetic lethal target of metformin in HCC. Mechanistically, metformin promotes DOCK1 phosphorylation, which activates RAC1 to facilitate cell survival, leading to metformin resistance. The DOCK1-selective inhibitor, TBOPP, potentiates anti-tumor activity by metformin in vitro in liver cancer cell lines and patient-derived HCC organoids, and in vivo in xenografted liver cancer cells and immunocompetent mouse liver cancer models. Notably, metformin improves overall survival of HCC patients with low DOCK1 levels but not among patients with high DOCK1 expression. This study shows that metformin effectiveness depends on DOCK1 levels and that combining metformin with DOCK1 inhibition may provide a promising personalized therapeutic strategy for metformin-resistant HCC patients.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13238-022-00906-6.