A deep learning-based hybrid model of global terrestrial evaporation.
Ontology highlight
ABSTRACT:
SUBMITTER: Koppa A
PROVIDER: S-EPMC8993934 | biostudies-literature | 2022 Apr
REPOSITORIES: biostudies-literature
Publications
A deep learning-based hybrid model of global terrestrial evaporation.
Koppa Akash A Rains Dominik D Hulsman Petra P Poyatos Rafael R Miralles Diego G DG
Nature communications 20220408 1
Terrestrial evaporation (E) is a key climatic variable that is controlled by a plethora of environmental factors. The constraints that modulate the evaporation from plant leaves (or transpiration, E<sub>t</sub>) are particularly complex, yet are often assumed to interact linearly in global models due to our limited knowledge based on local studies. Here, we train deep learning algorithms using eddy covariance and sap flow data together with satellite observations, aiming to model transpiration s ...[more]
Similar Datasets
Project description:Many epidemics have afflicted humanity throughout history, claiming many lives. It has been noted in our time that heart disease is one of the deadliest diseases that humanity has confronted in the contemporary period. The proliferation of poor habits such as smoking, overeating, and lack of physical activity has contributed to the rise in heart disease. The killing feature of heart disease, which has earned it the moniker the "silent killer," is that it frequently has no apparent signs in advance. As a result, research is required to develop a promising model for the early identification of heart disease using simple data and symptoms. The paper's aim is to propose a deep stacking ensemble model to enhance the performance of the prediction of heart disease. The proposed ensemble model integrates two optimized and pre-trained hybrid deep learning models with the Support Vector Machine (SVM) as the meta-learner model. The first hybrid model is Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM) (CNN-LSTM), which integrates CNN and LSTM. The second hybrid model is CNN-GRU, which integrates CNN with a Gated Recurrent Unit (GRU). Recursive Feature Elimination (RFE) is also used for the feature selection optimization process. The proposed model has been optimized and tested using two different heart disease datasets. The proposed ensemble is compared with five machine learning models including Logistic Regression (LR), Random Forest (RF), K-Nearest Neighbors (K-NN), Decision Tree (DT), Naïve Bayes (NB), and hybrid models. In addition, optimization techniques are used to optimize ML, DL, and the proposed models. The results obtained by the proposed model achieved the highest performance using the full feature set.
| S-EPMC9777370 | biostudies-literature
Project description:Deep learning, a subset of artificial intelligence, gives easy way for the analytical and physical tasks to be done automatically. There is a less necessity for human intervention while performing these tasks. Deep hybrid learning is a blended approach to combine machine learning with deep learning. A hybrid deep learning (HDL) model using convolutional neural network (CNN), residual network (ResNet) and long short term memory (LSTM) is proposed for better course selection of the enrolled candidates in an online learning platform. In this work, a hybrid framework that facilitates the analysis and design of a recommendation system for course selection is developed. A student's schedule for the next course should consist of classes in which the student has shown interest. For universities to schedule classes optimally, they need to know what courses each student wants to take before each course begins. The proposed recommendation system selects the most appropriate course that can encourage students to base their selection on informed decision making. This system will enable learners to obtain the correct choices of courses to be studied.
| S-EPMC10703022 | biostudies-literature
Project description:BackgroundThe zone adjacent to a transcription start site (TSS), namely, the promoter, is primarily involved in the process of DNA transcription initiation and regulation. As a result, proper promoter identification is critical for further understanding the mechanism of the networks controlling genomic regulation. A number of methodologies for the identification of promoters have been proposed. Nonetheless, due to the great heterogeneity existing in promoters, the results of these procedures are still unsatisfactory. In order to establish additional discriminative characteristics and properly recognize promoters, we developed the hybrid model for promoter identification (HMPI), a hybrid deep learning model that can characterize both the native sequences of promoters and the morphological outline of promoters at the same time. We developed the HMPI to combine a method called the PSFN (promoter sequence features network), which characterizes native promoter sequences and deduces sequence features, with a technique referred to as the DSPN (deep structural profiles network), which is specially structured to model the promoters in terms of their structural profile and to deduce their structural attributes.ResultsThe HMPI was applied to human, plant and Escherichia coli K-12 strain datasets, and the findings showed that the HMPI was successful at extracting the features of the promoter while greatly enhancing the promoter identification performance. In addition, after the improvements of synthetic sampling, transfer learning and label smoothing regularization, the improved HMPI models achieved good results in identifying subtypes of promoters on prokaryotic promoter datasets.ConclusionsThe results showed that the HMPI was successful at extracting the features of promoters while greatly enhancing the performance of identifying promoters on both eukaryotic and prokaryotic datasets, and the improved HMPI models are good at identifying subtypes of promoters on prokaryotic promoter datasets. The HMPI is additionally adaptable to different biological functional sequences, allowing for the addition of new features or models.
| S-EPMC9158169 | biostudies-literature
Project description:The regulation of metabolites comprises the fundamental biological processes of human. More new phenotypes of classic metabolites were revealed, while the devolution of functional targets and mechanisms of them still have many gaps due to diversity in structures and functions. Here, we report a bioinformation and attention mechanism based deep learning model leveraging abundant bioinformation of metabolites for functional target prediction of specific phenotypes. Taking advantage of this model, the potential target database of 3,382 common human metabolites were resourced publicly. Norepinephrine was identified as a selective peroxidase inhibitor of peroxiredoxin 1 (PRDX1) for atherosclerosis aggravation in model validation experiments. Salvianolic acid A (SAA) was screened out as PRDX1 exogenous stabilization compound for rescuing norepinephrine’s aggravation effects in vivo. Besides, bulk RNA-Sequencing analysis indicated norepinephrine aggravates atherosclerosis through activating transcriptional activity of SREBP2 in a ROS dependent manner. Overall, our research provides a new computational regime for metabolites target identification besides experimental chemical proteomics and PRDX1 stabilization was validated a potential druggable target for cardiovascular diseases intervention.
2025-05-07 | GSE282003 | GEO
Project description:Crafting a lucrative stock trading strategy is pivotal in the realm of investments. However, the task of devising such a strategy becomes challenging task the intricate and ever-changing situation of the stock market. In recent years, with the development of artificial intelligence (AI), some AI technologies have been proven to be successfully applied in stock price and asset management. For example, long short-term memory networks (LSTM) can be used for predicting stock price variation, reinforcement learning (RL) can be used for control stock trading, however, they are generally used separately and cannot achieve simultaneous prediction and trading. In this study, we propose a hybrid deep learning model to predict stock prices and control stock trading to manage assets. LSTM is responsible for predicting stock prices, while RL is responsible for stock trading based on the predicted price trends. Meanwhile, to reduce uncertainty in the stock market and maximize stock assets, the proposed LSTM model can predict the average directional index (ADX) to comprehend the stock trends in advance and we also propose several constraints to assist assets management, thereby reducing the risk and maximizing the stock assets. In our results, the hybrid model yields an average R 2 value of 0.94 when predicting price variations. Moreover, employing the proposed approach, which integrates ADX and constraints, the hybrid model augments stock assets to 1.05 times than initial assets.
| S-EPMC11639306 | biostudies-literature
Project description:BackgroundWe aimed to establish and validate a deep learning-based hybrid artificial intelligence (AI) model for the objective morphometric and colorimetric assessment of vitiligo lesions.MethodsTwo main datasets containing curated images of vitiligo lesions from Chinese patients (Fitzpatrick skin types III or IV) were established, including one with 2,720 images for lesion localization study and the other with 1,262 images for lesion segmentation study. Besides, an additional test set containing 145 images of vitiligo lesions from other Fitzpatrick skin types (I, II, or V) was also generated. A 3-stage hybrid model was constructed. YOLO v3 (You Only Look Once, v3) architecture was trained and validated to classify and localize vitiligo lesions, with sensitivity and error rate as primary performance outcomes. Then a segmentation study comparing 3 deep convolutional neural networks (DCNNs), Pyramid Scene Parsing Network (PSPNet), UNet, and UNet++, was carried out based on the Jaccard index (JI). The architecture with the best performance was integrated into the model. Three add-on metrics, namely VAreaA, VAreaR, and VColor were finally developed to measure absolute, relative size changes and pigmentation, respectively. Agreement between the AI model and dermatologist evaluators were assessed.ResultsThe sensitivity of the YOLO v3 architecture to detect vitiligo lesions was 92.91% with an error rate of 14.98%. The UNet++ architecture outperformed the others in the segmentation study (JI, 0.79) and was integrated into the model. On the additional test set, however, the model achieved a lower detection sensitivity (72.41%) and a lower segmentation score (JI, 0.69). With respect to size changes, no difference was observed between the AI model, trained dermatologists (W=0.812, P<0.05), and Photoshop analysis (P=0.075, P=0.212 respectively), which all displayed good concordance.ConclusionsWe developed a novel, convenient, objective, and quantitative deep learning-based hybrid model which simultaneously evaluated both morphometric and colorimetric vitiligo lesions from patients with Fitzpatrick skin types III or IV, rendering it suitable for the assessment of severity of vitiligo lesions in Asians in both clinic and research scenarios. More work is also warranted for its use in other ethnic skin groups.
| S-EPMC9201159 | biostudies-literature
Project description:BackgroundThe novel coronavirus pneumonia that began to spread in 2019 is still raging and has placed a burden on medical systems and governments in various countries. For policymaking and medical resource decisions, a good prediction model is necessary to monitor and evaluate the trends of the epidemic. We used a long short-term memory (LSTM) model and the improved hybrid model based on ensemble empirical mode decomposition (EEMD) to predict COVID-19 trends; Methods: The data were collected from the Harvard Dataverse. Epidemic data from 21 January 2020 to 25 April 2021 for California, the most severely affected state in the United States, were used to develop an LSTM model and an EEMD-LSTM hybrid model, which is an LSTM model combined with ensemble empirical mode decomposition. In this study, ninety percent of the data were adopted to fit the models as a training set, while the subsequent 10% were used to test the prediction effect of the models. The mean absolute percentage error, mean absolute error, and root mean square error were used to evaluate the prediction performances of the models; Results: The results indicated that the number of confirmed cases in California was increasing as of 25 April 2021, with no obvious evidence of a sharp decline. On 25 April 2021, the LSTM model predicted 3666418 confirmed cases, whereas the EEMD-LSTM predicted 3681150. The mean absolute percentage errors for the LSTM and EEMD-LSTM models were 0.0151 and 0.0051, respectively. The mean absolute and root mean square errors were 5.58 × 104 and 5.63 × 104 for the LSTM model and 1.9 × 104 and 2.43 × 104 for the EEMD-LSTM model, respectively; Conclusions: The results showed the advantage of an EEMD-LSTM model over a single LSTM model, and the established EEMD-LSTM model may be suitable for monitoring and evaluating the epidemic situation and providing quantitative analysis evidence for epidemic prevention and control.
| S-EPMC9819685 | biostudies-literature
Project description:A metagenome contains all DNA sequences from an environmental sample, including viruses, bacteria, archaea, and eukaryotes. Since viruses are of huge abundance and have caused vast mortality and morbidity to human society in history as a type of major pathogens, detecting viruses from metagenomes plays a crucial role in analyzing the viral component of samples and is the very first step for clinical diagnosis. However, detecting viral fragments directly from the metagenomes is still a tough issue because of the existence of a huge number of short sequences. In this study a hybrid Deep lEarning model for idenTifying vIral sequences fRom mEtagenomes (DETIRE) is proposed to solve the problem. First, the graph-based nucleotide sequence embedding strategy is utilized to enrich the expression of DNA sequences by training an embedding matrix. Then, the spatial and sequential features are extracted by trained CNN and BiLSTM networks, respectively, to enrich the features of short sequences. Finally, the two sets of features are weighted combined for the final decision. Trained by 220,000 sequences of 500 bp subsampled from the Virus and Host RefSeq genomes, DETIRE identifies more short viral sequences (<1,000 bp) than the three latest methods, such as DeepVirFinder, PPR-Meta, and CHEER. DETIRE is freely available at Github (https://github.com/crazyinter/DETIRE).
| S-EPMC10313334 | biostudies-literature
Project description:Oligonucleotide aptamers are typically identified through a rigorous and time-consuming process known as systematic evolution of ligands by exponential enrichment (SELEX), which requires 20 to 30 iterative rounds to eliminate non/weak binding sequences and enrich tight binding sequences with high affinity. Moreover, inherent experimental biases and non-specific interactions within SELEX could inadvertently exclude high-affinity candidates, leading to a high failure rate. To address these challenges, we proposed DeepAptamer for identifying high-affinity sequences from unenriched early SELEX rounds. As a hybrid neural network model combining convolutional neural networks and bidirectional long short-term memory, DeepAptamer integrated sequence composition and structural features to predict aptamer binding affinities and potential binding motifs. Trained on comprehensive SELEX data, DeepAptamer outperformed existing models in accuracy as substantiated by experimental evidence. More importantly, DeepAptamer effectively identified key nucleotides for target binding. DeepAptamer can efficiently identify high-affinity aptamers against various targets, enhancing its potential to discover promising sequences in initial screening stages and obviating the 20-30 iterative selection rounds required for full enrichment of selection pools. This represented a notable leap forward in aptamer technology, with broad implications for its application across a spectrum of selection targets.
| S-EPMC11787022 | biostudies-literature
Project description:BackgroundmicroRNAs (miRNAs) have been shown to play essential roles in a wide range of biological processes. Many computational methods have been developed to identify targets of miRNAs. However, the majority of these methods depend on pre-defined features that require considerable efforts and resources to compute and often prove suboptimal at predicting miRNA targets.ResultsWe developed a novel hybrid deep learning-based (DL-based) approach that is capable of predicting miRNA targets at a higher accuracy. This approach integrates convolutional neural networks (CNNs) that excel in learning spatial features and recurrent neural networks (RNNs) that discern sequential features. Therefore, our approach has the advantages of learning both the intrinsic spatial and sequential features of miRNA:target. The inputs for our approach are raw sequences of miRNAs and genes that can be obtained effortlessly. We applied our approach on two human datasets from recently miRNA target prediction studies and trained two models. We demonstrated that the two models consistently outperform the previous methods according to evaluation metrics on test datasets. Comparing our approach with currently available alternatives on independent datasets shows that our approach delivers substantial improvements in performance. We also show with multiple evidences that our approach is more robust than other methods on small datasets. Our study is the first study to perform comparisons across multiple existing DL-based approaches on miRNA target prediction. Furthermore, we examined the contribution of a Max pooling layer in between the CNN and RNN and demonstrated that it improves the performance of all our models. Finally, a unified model was developed that is robust on fitting different input datasets.ConclusionsWe present a new DL-based approach for predicting miRNA targets and demonstrate that our approach outperforms the current alternatives. We supplied an easy-to-use tool, miTAR, at https://github.com/tjgu/miTAR . Furthermore, our analysis results support that Max Pooling generally benefits the hybrid models and potentially prevents overfitting for hybrid models.
| S-EPMC7912887 | biostudies-literature