The prediction of drug-target affinity (DTA) plays an increasingly important role in drug discovery. Nowadays, lots of prediction methods focus on feature encoding of drugs and proteins, but ignore the importance of feature aggregation. However, the increasingly complex encoder networks lead to the loss of implicit information and excessive model size. To this end, we propose a deep-learning-based approach namely FusionDTA. For the loss of implicit information, a novel muti-head linear attention mechanism was utilized to replace the rough pooling method. This allows FusionDTA aggregates global information based on attention weights, instead of selecting the largest one as max-pooling does. To solve the redundancy issue of parameters, we applied knowledge distillation in FusionDTA by transfering learnable information from teacher model to student. Results show that FusionDTA performs better than existing models for the test domain on all evaluation metrics. We obtained concordance index (CI) index of 0.913 and 0.906 in Davis and KIBA dataset respectively, compared with 0.893 and 0.891 of previous state-of-art model. Under the cold-start constrain, our model proved to be more robust and more effective with unseen inputs than baseline methods. In addition, the knowledge distillation did save half of the parameters of the model, with only 0.006 reduction in CI index. Even FusionDTA with half the parameters could easily exceed the baseline on all metrics. In general, our model has superior performance and improves the effect of drug–target interaction (DTI) prediction. The visualization of DTI can effectively help predict the binding region of proteins during structure-based drug design.
This study examined the Chabagou River watershed in the gully region of the Loess Plateau in China’s Shaanxi Province, and was based on measured precipitation and runoff data in the basin over a 52-year period (1959–2010), land-use types, normalized difference vegetation index (NDVI), and other data. Statistical models and distributed hydrological models were used to explore the influences of climate change and human activity on the hydrological response and on the temporal and spatial evolution of the basin. It was found that precipitation and runoff in the gully region presented a downward trend during the 52-year period. Since the 1970s, the hydrological response to human activities has become the main source of regional hydrological evolution. Evapotranspiration from the large silt dam in the study area has increased. The depth of soil water decreased at first, then it increased by amount that exceeded the evaporation increase observed in the second and third change periods. The water and soil conservation measures had a beneficial effect on the ecology of the watershed. These results provide a reference for water resource management and soil and water conservation in the study area.
Abstract. As the birthplace of Yangtze River, the Yellow River and Lancang Rive, Source Region of Three Rivers (SRTR) is an important resource for fresh water supplement in China. SRTR also has very obvious ecological function which forms ecological security barrier for China's Qinghai-Tibet plateau. The inland lakes here play an important role for the water cycle in the plateau. The monitoring results were extracted with TM data from 1976 to 2009. The results show that from 1976 to 2009 the lakes' area in SRTR dropped first and then expanded with 2000 as sector. The lakes area was 6778 km2 in 2009, about 1.90% of the whole region, and increased than 1976 by 133.15 km2. Most of the large lakes above 80 km2 have the same change trend. The expanded lakes increased in number gradually, while the changes in the amplitude and time characteristics were different. From 1976 to 2000, the number of new lakes increased while died lakes dropped; and from 2000 to 2009 it is just on the contrary. In the study the index of lake change trend (ILCT) was adopted to contrast lake atrophy condition. With ILCT 24.55 there is an expansion trend for the lakes in SRTR during the last 35 years. The lakes with ILCT's absolute value greater than 1 were those merged with or disconnected from surrounding smaller lakes. Here the precipitation and snow melt are main supplies for the lakes. The change of lakes' area has well correlated with precipitation, and weak correlated with temperature from 1976 to 2009. But from 2000 to 2009, there has a strong correlation with precipitation, temperature. All these show from the side that the precipitation and snow melt are important factors to influence the lakes’ change. The lakes have the coordination function for the good ecological environment in the region. The conclusions from the study can provide references in response to climate change research and rational utilization of water resources in SRTR.
Through the finite element software ABAQUS, the finite element model considering the initial imperfection and residual stress is established, and the finite element results are compared with the collected test results to verify the reliability of the numerical model. By analyzing the ultimate carrying capacity of I section of axial compression with different aspect ratios, the design method of ultimate carrying capacity of axial compression members of hot rolled I section from thick to thin is studied. The result of Overall Interaction Concept (OIC) for hot rolled I section steel under axial compression is obtained by using the finite element calculation results, and the results are compared with the Eurocode (EN1993-1-1) and the Chinese steel structure design standard (GB50017-2017), so as to study the accuracy of the recommend design method. Results found that: i) the calculation result from EC3 of the cross section classification concept most conservative or unsafe, ii) the results from GB almost all conservative, iii) comparing with the existed design methods the OIC design method reflect the relationship between carrying capacity and the the generalized relative slenderness, that can accurately predict ultimate carrying capacity. Research shows that OIC is a more effective and accurate method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.