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.     

AvrRps4 effector family processing and recognition in lettuce

During pathogenesis, effector proteins are secreted from the pathogen to the host plant to provide virulence activity for invasion of the host. However, once the host plant recognizes one of the delivered effectors, effector‐triggered immunity activates a robust immune and hypersensitive response (HR). In planta, the effector AvrRps4 is processed into the N‐terminus (AvrRps4^N) and the C‐terminus (AvrRps4^C). AvrRps4^C is sufficient to trigger HR in turnip and activate AtRRS1/AtRPS4‐mediated immunity in Arabidopsis; on the other hand, AvrRps4^N induces HR in lettuce. Furthermore, AvrRps4^N‐mediated HR requires a conserved arginine at position 112 (R112), which is also important for full‐length AvrRps4 (AvrRps4^F) processing. Here, we show that effector processing and effector recognition in lettuce are uncoupled for the AvrRps4 family. In addition, we compared effector recognition by lettuce of AvrRps4 and its homologues, HopK1 and XopO. Interestingly, unlike for AvrRps4 and HopK1, mutation of the conserved R111 in XopO by itself was insufficient to abolish recognition. The combination of amino acid substitutions arginine 111 to leucine with glutamate 114 to lysine abolished the XopO‐mediated HR, suggesting that AvrRps4 family members have distinct structural requirements for perception by lettuce. Together, our results provide an insight into the processing and recognition of AvrRps4 and its homologues.     

不同血清型肉毒毒素受体结合域研究进展

肉毒毒素(botulinum neurotoxin, BoNT)是人类已知毒性最强的蛋白质之一,可以引起肌肉松弛麻痹,严重时可导致死亡。肉毒毒素共分为7种血清型(BoNT/A-BoNT/G),根据氨基酸序列差异可进一步分为40多种亚型。肉毒毒素分子结构由3个基本结构域组成：重链羧基端细胞受体结合域、氨基端的易位域和轻链催化域。在运动神经元表面,受体结合域首先与聚唾液酸神经节苷脂结合,随后与突触囊泡蛋白2或突触囊泡结合蛋白结合形成双受体复合物。每种血清型的受体结合域都必须与其相应受体结合才能发挥作用。肉毒毒素的结构功能及其对宿主的作用一直都是研究热点。近年来,因受体结合域可以促进肉毒毒素与运动神经元膜特异性结合,而成为新的研究方向。本综述将概述不同血清型肉毒毒素与受体结合过程中受体结合域结构变化和结合位点差异。通过分析不同血清型及亚型的序列以及受体结合域结构特征,可以更好地了解细胞受体结合域的序列差异和功能,并为肉毒毒素的治疗策略提供新思路。     

云南主要有害实蝇种类及区划

采用诱剂诱捕辅以受害瓜果调查的方法,系统梳理了云南有害实蝇类群,确定主要有害种类25种。基于云南自然地理特征,采用聚类分析法深入探讨了云南实蝇区系,提出云南实蝇地理区划可分为德宏、版纳盆谷区,南部边缘中低山宽谷区,滇西南山原峡谷区,中东部高原、岩溶山原区以及北部高原山地区等5大区域。研究发现,受热量条件制约,云南主要有害实蝇类群呈南北向梯度递减;受纵向岭谷区"通道-阻隔"综合作用影响,云南西南部河谷地区是实蝇危害较为严重的区域。首次系统揭示了云南主要有害实蝇的地理分布规律。     

水土保持林土壤改良效益评价研究

森林效益评价是目前林业研究的热点,也是难点问题,通过评价水土保持林土壤改良效益,对效益评价的方法和思路进行了探索。指出水土保持林土壤改良效益指标体系的总目标层为效益评价综合指数,准则层包括土壤肥力、抗蚀性能和抗冲性能３个方面,指标层由水稳性团聚体含量、有机质含量、土壤酶活性、ｐＨ值、平均重量直径、团聚度、分散率、土壤硬度和渗透系数构成。在计算效益评价综合指数时,首先要对各项指标的实测数据进行处理,     

Depression compromises antiviral innate immunity via the AVP-AHI1-Tyk2 axis

Depression is a serious public-health issue. Recent reports have suggested higher susceptibility to viral infections in depressive patients. However, how depression affects antiviral innate immune signaling remains unknown. Here, we revealed a reduction in expression of Abelson helper integration site 1 (AHI1) in the peripheral blood mononuclear cells (PBMCs) and macrophages from the patients with major depressive disorder (MDD), which leads to attenuated antiviral immune response. We found that depression-related arginine vasopressin (AVP) induces reduction of AHI1 in macrophages. Further studies demonstrated that AHI1 is a critical stabilizer of basal type-I-interferon (IFN-I) signaling. Mechanistically, AHI1 recruits OTUD1 to deubiquitinate and stabilize Tyk2, while AHI1 reduction downregulates Tyk2 and IFN-I signaling activity in macrophages from both MDD patients and depression model mice. Interestingly, we identified a clinical analgesic meptazinol that effectively stimulates AHI1 expression, thus enhancing IFN-I antiviral defense in depression model mice. Our study promotes the understanding of the signaling mechanisms of depression-mediated antiviral immune dysfunction, and reveals meptazinol as an enhancer of antiviral innate immunity in depressive patients.Subject terms: Innate immunity, Ubiquitylation, Cell signalling     

三种回归分析方法在Hyperion影像LAI反演中的比较

借助GPS进行地面精确定位,利用LAI-2000冠层分析仅在攸县黄丰桥林场开展130个样地(60m×60m)的叶面积指数(Leaf Area Index,LAI)测量.采用FLAASH模块对Hyperion数据进行大气校正并与地面同步冠层观测数据进行拟合,通过研究地面实测LAI与Hyperion影像波段及其衍生的系列植被指数(NDVI、RVI等)的相关性,筛选出估算叶面积指数的植被指数因子.应用曲线估计、逐步回归及偏最小二乘三种回归分析技术分别建立叶面积指数的最优估算模型.结果表明:参与建模的因子中,比值植被指数(RVI)与LAI的相关性最大,敏感性最高,其次是SARVI0.1,NDVI705,NDVI,SARVI0.1,SARVI0.25;曲线估计、逐步回归分析和偏最小二乘回归三种分析方法所建的6个回归模型中,偏最小二乘回归的拟合效果最好,预测值与实测值的决定系数R2为0.84、曲线估计的拟合效果最低,预测值与实测值的决定系数R2为0.64;建模精度分析表明,选用5-6个自变量因子进行LAI建模是可靠的,以6个植被因子建立的偏最小二乘回归模型预测精度最高.     

一株高效广谱染料降解细菌的分离鉴定及其脱色特性初探

从土壤样品中分离到一株高效染料脱色菌株N-4,根据形态学特征及16S rDNA基因序列分析,该菌株初步鉴定为Leucobacter sp.。利用表面响应法(RSM)对菌株N-4脱色活性深蓝K-R的主要因素进行优化,实验结果表明,菌株N-4脱色K-R的最优条件为:湿菌量10 g/L,染料浓度222 mg/L,硫酸铵1.5 g/L,果糖3.5 g/L,最佳脱色率为100%。此外,实验证明其对多种染料均具有较高的脱色效率。同时,考察了金属离子对染料脱色效率的影响,其中K+、Ca2+、Mg2+、Ba2+、Mn2+等对脱色具有促进作用,而Ni2+、Cu2+、Hg2+对脱色具有明显的抑制作用。     

滇池外海蓝藻水华爆发反演及规律探讨

气象条件和营养盐浓度一直被认为是导致蓝藻水华爆发的两个重要因素。通过滇池外海Chla浓度时空分异性分析,得出晖湾中测点最易爆发蓝藻水华且爆发时间集中在每年6—9月。同时,采用基于缺失数据多重插补的EMB算法将气象条件和蓝藻水华爆发的不完全数据集进行反演,建立了滇池外海2004—2008年4—10月完整的气象、营养盐及蓝藻爆发的基础数据集,解决了表观蓝藻水华爆发研究中观测数据缺失的问题。据此,探讨了滇池外海晖湾中测点Chla、TN和TP与蓝藻水华爆发关系,进而提出了控制滇池外海蓝藻水华的一种新思路。     

叶脉结构与功能及其对叶片经济谱的影响

叶脉由贯穿于叶肉内部的维管组织及其外围机械组织构成,多样化的脉序及网络结构使叶脉系统发生变异和功能分化。该文综述了叶脉系统结构与功能的最新研究进展。通过聚焦叶脉分级系统的结构与功能及其在叶片经济谱(LES)中的重要性,解释叶脉性状与其它叶片功能性状之间的关系及机制。不同等级叶脉在机械支撑与水分运输方面存在功能分化,其中1–3级粗脉在维持叶片形状和叶表面积以及物理支撑方面发挥重要作用,有利于维持叶片最大受光面积;4级及以上细脉具有水分调节功能,它们与气孔相互协调,影响叶片水分运输、蒸腾散热和光合作用速率。叶片生长过程与叶脉发育的动态变化模式决定叶脉密度,并影响叶脉密度与叶片大小之间的关系:叶面积与粗脉密度呈显著负相关,与粗脉直径呈显著正相关,而与细脉密度无关。与叶脉性状相关的叶片经济谱框架模型预测,叶脉密度较高的叶片寿命短、比叶重较小,叶片最大碳同化速率、代谢速率以及资源获取策略潜力较高。