Sequence Dependence Research Articles

Label-Free and Sequence-Independent Isothermal Amplification Strategy for the Simultaneous Detection of Genomic 5-Methylcytosine and 5-Hydroxymethylcytosine.

5-Methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC) are crucial epigenetic modifications in eukaryotic genomic DNA that regulate gene expression and are associated with the occurrence of various cancers. Here, we combined bisulfite conversion with 4-acetamido-2,2,6,6-tetramethyl-1-oxopiperridinium tetrafluoroborate (ACT+BF4-, TCI) oxidation to develop a label-free and sequence-independent isothermal amplification (BTIA) assay for a genome-wide 5mC and 5hmC analysis. The BTIA strategy can distinguish 5mC and 5hmC signatures from other bases with high sensitivity and good specificity, avoiding sophisticated chemical modifications and expensive protein labeling. Moreover, the utilization of terminal deoxynucleotidyl transferase (TdT) enables the proposed strategy to detect global 5mC and 5hmC without sequence dependence. With only 78 ng of input of genomic DNA, global 5mC and 5hmC levels were accurately quantified in cells (including cancer cells of A549, T47D, and K562 and normal cells of HEK-293T, CHO, and NRK-52E) and clinical whole blood samples (including healthy control, precancerous cervical cancer, and confirmed cervical cancer) within 18 h. The detection results suggested that 5mC was highly expressed in cancer cells. More importantly, a significant increase in 5mC was observed in precancerous cervical cancer and further upregulation in confirmed cervical cancer, suggesting a correlation between 5mC and cancer occurrence and development. However, 5hmC showed the reverse result in these tested cells and clinical samples. Collectively, the BTIA strategy can be easily performed on the ordinary heating apparatus in almost all research and medical laboratories, showing a significant application in the early screening of cervical cancer in the clinic.

Customer churn prediction model based on hybrid neural networks

In today’s competitive market environment, accurately identifying potential churn customers and taking effective retention measures are crucial for improving customer retention and ensuring the sustainable development of an organization. However, traditional machine learning algorithms and single deep learning models have limitations in extracting complex nonlinear and time-series features, resulting in unsatisfactory prediction results. To address this problem, this study proposes a hybrid neural network-based customer churn prediction model, CCP-Net. In the data preprocessing stage, the ADASYN sampling algorithm balances the sample sizes of churned and non-churned customers to eliminate the negative impact of sample imbalance on the model performance. In the feature extraction stage, CCP-Net uses Multi-Head Self-Attention to learn the global dependencies of the input sequences, combines with BiLSTM to capture the long-term dependencies in the sequential data, and uses CNN to extract the local features, and ultimately generates the prediction results. Experimental results of cross-validation on Telecom, Bank, Insurance, and News datasets show that CCP-Net outperforms the comparison algorithms in all performance metrics. For example, CCP-Net achieves a Precision of 92.19% on the Telecom dataset, 91.96% on the Bank dataset, 95.87% on the Insurance dataset, and 95.12% on the News dataset, which compares to other hybrid neural network models, the performance improvement of CCP-Net ranges from 1% to 3%. These results indicate that the design of the CCP-Net model effectively improves the accuracy and robustness of churn prediction, enabling it to be widely applied to different industries, especially in the financial, telecommunication, and media fields, to provide more comprehensive and effective churn management strategies for enterprises.

DYNAMIC GENERATIVE RESIDUAL GRAPH CONVOLUTIONAL NEURAL NETWORKS FOR ELECTRICITY THEFT DETECTION

Illegal electricity users pose a significant threat to the economic and security aspects of the power system by illicitly accessing or manipulating electrical resources. With the widespread adoption of Advanced Metering Infrastructure (AMI), researchers have turned to leveraging smart meter data for electricity theft detection. However, existing models rely on methods that model a single electricity load curve and cannot capture the temporal dependencies, periodicity, and underlying features between electricity consumption cycles. This study introduces a novel electricity theft detection method based on dynamic residual graph networks. Innovatively, it proposes a dynamic topological graph construction technique that allows for the real-time updating of adjacency matrices during the training process, thereby effectively capturing the complex relationships in electricity usage patterns. Utilizing the MixHop graph convolutional network, it delves into the temporal sequence dependencies, periodicity, and hidden characteristics within user electricity consumption data. Additionally, to address the issue of model instability caused by scarce theft data, we employ the SMOTE (Synthetic Minority Over-sampling Technique) oversampling technique and enhance overall classification performance by modifying class weights in the loss function. We trained this network architecture on the real SGCC (State Grid Corporation of China) dataset, and the results demonstrate its superiority over other mainstream existing models.

EXTRACTING EMOTION-CAUSE PAIRS: A BILSTM-DRIVEN METHODOLOGY

Emotions are fundamental to human interactions, intricately influencing communication, behavior, and perception. Emotion-Cause Pair Extraction (ECPE) is a critical task in natural language processing that identifies clause pairs associating emotions with their corresponding triggers within textual documents. Unlike traditional Emotion Cause Extraction (ECE), which relies on pre-annotated emotion clauses, our study introduces a novel end-to-end model for ECPE. This innovative approach utilizes the extensive NTCIR-13 English Corpus to establish a robust baseline for ECPE in English, showcasing significant performance improvements over conventional multi-stage methods. Central to our model is the incorporation of Bidirectional Long Short-Term Memory (BiLSTM) networks, enhancing the ability to capture both local and global dependencies in textual sequences. By effectively combining contextual and positional embeddings, our model accurately predicts emotion-cause relationships, paving the way for a deeper understanding of emotional dynamics in conversational contexts and facilitating causal inference. Furthermore, our research highlights superior performance metrics, aligning its efficacy with state-of-the-art techniques in the field. This study advances emotion recognition in natural language processing, providing valuable insights for nuanced analyses of human emotions within textual data. Additionally, our findings enhance understanding of emotional intelligence in user interaction modeling and conversational AI applications. Through the public availability of our dataset and model, we aim to foster collaboration and further research in this vital area, ultimately improving the capacity for emotional understanding in applications ranging from sentiment analysis to interactive learning.

Dual-Path Beat Tracking: Combining Temporal Convolutional Networks and Transformers in Parallel

The Transformer, a deep learning architecture, has shown exceptional adaptability across fields, including music information retrieval (MIR). Transformers excel at capturing global, long-range dependencies in sequences, which is valuable for tracking rhythmic patterns over time. Temporal Convolutional Networks (TCNs), with their dilated convolutions, are effective at processing local, temporal patterns with reduced complexity. Combining these complementary characteristics, global sequence modeling from Transformers and local temporal detail from TCNs enhances beat tracking while reducing the model’s overall complexity. To capture beat intervals of varying lengths and ensure optimal alignment of beat predictions, the model employs a Dynamic Bayesian Network (DBN), followed by Viterbi decoding for effective post-processing. This system is evaluated across diverse public datasets spanning various music genres and styles, achieving performance on par with current state-of-the-art methods yet with fewer trainable parameters. Additionally, we also explore the interpretability of the model using Grad-CAM to visualize the model’s learned features, offering insights into how the TCN-Transformer hybrid captures rhythmic patterns in the data.

Sequence-Specific Free Energy Changes in DNA/RNA Induced by a Single LNA-T Modification in Antisense Oligonucleotides.

2',4'-methylene bridged nucleic acid/locked nucleic acid (2',4'-BNA/LNA; LNA) is a modified nucleic acid that improves the function of antisense oligonucleotide therapeutics. In particular, LNA in the DNA strand increases its binding affinity for the target RNA. Predicting the binding affinities of LNA-containing antisense oligonucleotides and RNA duplexes is useful for designing antisense oligonucleotides. The nearest neighbor parameters may be useful for binding affinity prediction, similar to those for natural nucleic acids. However, the sequence dependence of the thermodynamic stability of DNA/RNA duplexes containing LNA remains unexplored. Therefore, in this study, we evaluated the thermodynamic stabilities of DNA/RNA duplexes containing a single LNA modification in the DNA strand. We found that LNA-stabilized DNA/RNA duplexes averaged -1.5 kcal mol-1. Our findings suggest that the thermodynamic stabilization effect of LNA is sequence-specific.

Adaptive Deep Reinforcement Learning Energy Management for Hybrid Electric Vehicles Considering Driving Condition Recognition

Adaptive energy management strategy (EMS) can adjust the underlying control scheme of hybrid electric vehicles (HEVs) according to different external conditions, thus maintaining high performance consistently. For this reason, this study designs a double-layer EMS for a power-split HEV, which combines deep reinforcement learning (DRL) with driving condition recognition (DCR) to achieve efficient operation of powertrain under different driving conditions. At the upper layer, the long-term dependencies in velocity sequences are captured through the sequential networks to enhance the accuracy of DCR. At the lower layer, multi-objective reward functions are formulated, and the optimal weight parameters are identified for different driving conditions. Furthermore, the deep deterministic policy gradient (DDPG) algorithm is employed to dynamically optimize power flow. The simulation experiments are conducted to verify the feasibility and the energy-saving effects of the proposed strategy. The results indicate that integrating variations in driving conditions into the energy allocation of HEVs can further improve the fuel economy and maintain battery health. Specifically, the method enhances the fuel economy by an average of 10.76%, compared to a number of traditional benchmark strategies.

Application of state of health estimation and remaining useful life prediction for lithium-ion batteries based on AT-CNN-BiLSTM

Ensuring the long-term safe usage of lithium-ion batteries hinges on accurately estimating the State of Health \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$(\ extrm{SOH})$$\\end{document} and predicting the Remaining Useful Life (RUL). This study proposes a novel prediction method based on a AT-CNN-BiLSTM architecture. Initially, key parameters such as voltage, current, temperature, and SOH are extracted and averaged for each cycle to ensure the uniformity and reliability of the input data. The CNN is utilized to extract deep features from the data, followed by BiLSTM to analyze the temporal dependencies in the data sequences. Since multidimensional parameter data are used to predict the SOH trend of lithium-ion batteries, an attention mechanism is employed to enhance the weight of highly relevant vectors, improving the model’s analytical capabilities. Experimental results demonstrate that the CNN-BiLSTM-Attention model achieves an absolute error of 0 in RUL prediction, an \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$R^{2}$$\\end{document} value greater than 0.9910 , and a MAPE value less than 0.9003 . Comparative analysis with hybrid neural network algorithms such as LSTM, BiLSTM, and CNN-LSTM confirms the proposed model’s high accuracy and stability in SOH estimation and RUL prediction.

TCM: An efficient lightweight MLP-based network with affine transformation for long-term time series forecasting

Time series forecasting (TSF) involves extracting underlying patterns from past information to predict future sequences over a specific period. Extending the prediction length of time series and improving the prediction accuracy have always been challenging tasks. Autoregressive prediction methods based on Markov chains tend to accumulate errors over time. Although Transformer-based methods with various self-attention mechanisms have shown some improvements, they require higher memory and computational resources. In this work, we present an effective MLP-based TSF framework named TCM, which models the sequence and channel dependencies separately using Token MLP and Channel MLP. Additionally, we employ the Affine Transformation to replace layer normalization or batch normalization, leading to substantial enhancements in both accuracy and inference speed. Compared to current state-of-the-art long-term time series forecasting models, TCM achieves 6.0% relative improvement on seven real-world datasets, including electricity, weather, and illness domains. The TCM model, characterized by its efficiency and lightweight architecture, also makes it suitable for scenarios with high real-time requirements.

GRMD: A Two-Stage Design Space Exploration Strategy for Customized RNN Accelerators

Recurrent neural networks (RNNs) have produced significant results in many fields, such as natural language processing and speech recognition. Owing to their computational complexity and sequence dependencies, RNNs need to be deployed on customized hardware accelerators to satisfy performance and energy-efficiency constraints. However, designing hardware accelerators for RNNs is challenged by the vast design space and the reliance on ineffective optimization. An efficient automated design space exploration (DSE) strategy that can balance conflicting objectives is wanted. To address the low efficiency and insufficient universality of the resource allocation process employed for hardware accelerators, we propose an automated two-stage design space exploration (DSE) strategy for customized RNN accelerators. The strategy combines a genetic algorithm (GA) and a reinforcement learning (RL) algorithm, and it utilizes symmetrical exploration and exploitation to find the optimal solutions. In the first stage, the area of the hardware accelerator is taken as the optimization objective, and the GA is used for partial exploration purposes to narrow the design space while maintaining diversity. Then, the latency and power of the hardware accelerator are taken as the optimization objectives, and the RL algorithm is used in the second stage to find the corresponding Pareto solutions. To verify the effectiveness of the developed strategy, it is compared with other algorithms. We use three different network models as benchmarks: a vanilla RNN, LSTM, and a GRU. The results demonstrate that the strategy proposed in this paper can provide better solutions and can achieve latency, power, and area reductions of 9.35%, 5.34%, and 11.95%, respectively. The HV of GRMD is reduced by averages of 6.33%, 6.32%, and 0.67%, and the runtime is reduced by averages of 18.11%, 14.94%, and 10.28%, respectively. Additionally, given different weights, it can make reasonable trade-offs between multiple objectives.

Multi-Scale Adaptive Skeleton Transformer for action recognition

Transformer has demonstrated remarkable performance in various computer vision tasks. However, its potential is not fully explored in skeleton-based action recognition. On one hand, existing methods primarily utilize fixed function or pre-learned matrix to encode position information, while overlooking the sample-specific position information. On the other hand, these approaches focus on single-scale spatial relationships, while neglecting the discriminative fine-grained and coarse-grained spatial features. To address these issues, we propose a Multi-Scale Adaptive Skeleton Transformer (MSAST), including Adaptive Skeleton Position Encoding Module (ASPEM), Multi-Scale Embedding Module (MSEM), and Adaptive Relative Location Module (ARLM). ASPEM decouples spatial–temporal information in the position encoding procedure, which acquires inherent dependencies of skeleton sequences. ASPEM is also designed to be dependent on input tokens, which can learn sample-specific position information. The MSEM employs multi-scale pooling to generate multi-scale tokens that contain multi-grained features. Then, the spatial transformer captures multi-scale relations to address the subtle differences between various actions. Another contribution of this paper is that ARLM is presented to mine suitable location information for better recognition performance. Extensive experiments conducted on three benchmark datasets demonstrate that the proposed model achieves Top-1 accuracy of 94.9%/97.5% on NTU-60 C-Sub/C-View, 88.7%/91.6% on NTU-120 X-Sub/X-Set and 97.4% on NW-UCLA, respectively.

Reconstructing Richtmyer–Meshkov instabilities from noisy radiographs using low dimensional features and attention-based neural networks

We develop an ML-based approach for density reconstruction based on transformer neural networks. This approach is demonstrated in the setting of ICF-like double shell hydrodynamic simulations wherein the parameters related to material properties and initial conditions are varied. The new method can robustly recover the complex topologies given by the Richtmyer-Meshkoff instability (RMI) from a sequence of hydrodynamic features derived from radiographic images corrupted with blur, scatter, and noise. A noise model is developed to characterize errors in extracting features from synthetic radiographs of the simulated density field. The key component of the network is a transformer encoder that acts on a sequence of features extracted from noisy radiographs. This encoder includes numerous self-attention layers that act to learn temporal dependencies in the input sequences and increase the expressiveness of the model. This approach is shown to exhibit an excellent ability to accurately recover the RMI growth rates, despite the gas-metal interface being greatly obscured by radiographic noise. Our approach can be applied in a broad array of fields involving shock physics and material science.

Dynamic inferences of crash risks in freeway merging zones: a spatio-temporal deep learning model

Freeway merging zones are critical for freeway operations and management due to potential crashes arising from complex vehicle merging behaviours. This paper investigates the application of a spatio-temporal deep learning model to infer crash risks in the zones. We first introduce a crash risk index based on Time-To-Collision and vehicle merging patterns. An innovation is that the developed spatio-temporal transformer model can analyze the evolving risk index. This model effectively captures dynamic risk features through a multi-head attention mechanism within its spatio-temporal learning components. Numerical experiments on nine inference tasks with varying spatial resolutions show improved performance of the model with lower resolution. Moreover, the ST-Transformer model is benchmarked against three advanced deep learning models, which consistently demonstrates its superiority in capturing spatio-temporal dependence in risk sequences. This investigation significantly contributes to a richer understanding of proactive traffic safety, providing valuable insights for advanced freeway management and driver assistance systems.

Abstract B053: NucAE: Autoencoder-based enhancement of nucleosome occupancy signals from low-coverage cfDNA sequencing

Abstract Introduction: Cell-free DNA (cfDNA) is emerging as a valuable cancer biomarker, enabling both the detection and epigenetic characterization of malignancies through sequencing. cfDNA sequencing can reveal cancer-specific nucleosome occupancy patterns, which provide insights into tumor type and differentiation status, which define prognosis and therapy recommendations. Low-coverage whole-genome cfDNA sequencing is a cost-efficient approach to assay multiple cfDNA samples, however, the low coverage limits its sensitivity. We developed NucAE, a denoising autoencoder model, to enhance nucleosome occupancy signals from low-coverage cfDNA-sequencing. Methods: We designed NucAE to estimate nucleosome occupancy genome- wide from low-coverage sequencing data. To train the model, we undersampled high-coverage cfDNA-sequencing data and produced high- and low-coverage sample pairs at varying coverages (1x-90x). Inputs to NucAE include cfDNA fragment data and the corresponding genomic nucleotide sequence, with the model leveraging a symmetrical architecture comprising two convolutional layers in both the encoder and decoder, connected by ReLU activations. Mean squared error relative to the high-coverage signal was employed as the loss function. To evaluate the model independently, we performed peak detection from the nucleosome occupancy signals. We conducted a grid search to optimize hyperparameters. Results: We found that NucAE enhances the signal to noise ratio by detecting periodicities in nucleosome occupancy data. Denoising was improved by supplementing the cfDNA fragment signals with the genomic nucleotide sequences as input, which demonstrates that the model learned the nucleotide- sequence dependence of nucleosome positioning. Independent validation through peak detection on previously unseen samples demonstrated a significant improvement (p<0.001) in accuracy compared to raw signals across all tested coverages (1x-45x). Remarkably, at extremely low coverages (1x-2x), denoised signals allowed for more accurate peak detection than raw signals from 10-20x deeper sequencing. Conclusions: By using advanced machine-learning, we were able to enhance low-coverage cfDNA sequencing data to assay nucleosome positions at a nucleotide resolution. NucAE achieves an order of magnitude improvement over the low- coverage signals and greatly improves the utility of cfDNA-sequencing data to identify cancer- specific nucleosome occupancy, while preserving its low cost. Our model is the first to use autoencoders for genome-wide signal enhancement with potential use cases in many other areas of genomics. Citation Format: Zsolt Balázs, Jean Radig, Noah Wolford, Manuel Schürch, Michael Krauthammer. NucAE: Autoencoder-based enhancement of nucleosome occupancy signals from low-coverage cfDNA sequencing [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr B053.

CTFNet: Long-Sequence Time-Series Forecasting Based on Convolution and Time-Frequency Analysis.

Although current time-series forecasting methods have significantly improved the state-of-the-art (SOTA) results for long-sequence time-series forecasting (LSTF), they still have difficulty in capturing and extracting the features and dependencies of long-term sequences and suffer from information utilization bottlenecks and high-computational complexity. To address these issues, a lightweight single-hidden layer feedforward neural network (SLFN) combining convolution mapping and time-frequency decomposition called CTFNet is proposed with three distinctive characteristics. First, time-domain (TD) feature mining-in this article, a method for extracting the long-term correlation of horizontal TD features based on matrix factorization is proposed, which can effectively capture the interdependence among different sample points of a long time series. Second, multitask frequency-domain (FD) feature mining-this can effectively extract different frequency feature information of time-series data from the FD and minimize the loss of data features. Integrating multiscale dilated convolutions, simultaneously focusing on both global and local context feature dependencies at the sequence level, and mining the long-term dependencies of the multiscale frequency information and the spatial dependencies among the different scale frequency information, break the bottleneck of data utilization, and ensure the integrity of feature extraction. Third, highly efficient-the CTFNet model has a short training time and fast inference speed. Our empirical studies with nine benchmark datasets show that compared with state-of-the-art methods, CTFNet can reduce prediction error by 64.7% and 53.7% for multivariate and univariate time series, respectively.

Neoliberalism in question: The Philippines' nurse education and labour export as liberal neo‐statist development agenda

AbstractMany scholars have used neoliberalism as an analytical framework to examine the Philippines' labour export policy. While neoliberalism entails a retreat of the state in favour of market reforms, evidence shows that state intervention of the market becomes larger and stronger over time. This paper utilises liberal neo‐statism as an alternative framework to understand the Philippines' nurse labour export by explaining that the state's role is larger than and goes beyond labour brokerage. Following the historical institutionalism approach, we show the significant timing, sequence, and path dependence that affect the emergence of institutions that govern the Philippines' nurse labour export. Our paper reveals how specific policies and regulations in labour export are tucked within the disguise of market reforms, but which are manifest within a larger state's control. These policies serve as the state's apparatus for remittance generation and protection of migrant labour rights and welfare.

Ramachandran-like Conformational Space for DNA.

DNA's ability to exist in a wide variety of structural forms, subforms, and secondary motifs is fundamental to numerous biological processes and has driven the development of biotechnological applications. Major determinants of DNA flexibility are the multiple torsional degrees of freedom around the phosphodiester backbone. This high complexity can be rationalized by using two pseudotorsional angles linking atoms P and C4', from which Ramachandran-like plots can be built. In this contribution, we explore the distribution of η (eta: C4'i-1-Pi-C4'i-Pi+1) and θ (theta: Pi-C4'i-Pi+1-C4'i+1) angles in known experimental structures retrieved from the Protein Data Bank (PDB), subdividing the conformational space into different datasets. After the removal of the canonical/helical conformations typical of the B-form, we find the existence of a conformational map with clearly permitted and forbidden regions. Some of these regions are populated with specific DNA forms, like Z- or A-DNA, or by specific secondary motifs, like G-quadruplexes and junctions. We evaluated the sequence dependency and energy relationship among the high-density regions identified in the η-θ space. Furthermore, we analyzed the effect produced by proteins and cations when bound to DNA, finding that specific proteins produce some nonhelical conformations, while other regions appear to be stabilized by divalent cations.

RS-Net: Hyperspectral Image Land Cover Classification Based on Spectral Imager Combined with Random Forest Algorithm

Recursive neural networks and transformers have recently become dominant in hyperspectral (HS) image classification due to their ability to capture long-range dependencies in spectral sequences. Despite the success of these sequential architectures, mainstream deep learning methods primarily handle two-dimensional structured data. However, challenges such as the curse of dimensionality, spectral variability, and confounding factors in hyperspectral remote sensing images limit their effectiveness, especially in remote sensing applications. To address this issue, this paper proposes a novel land cover classification algorithm that integrates random forests with a spectral transformer network structure (RS-Net). Firstly, this paper presents a combination of the Gramian Angular Field (GASF) and Gramian Angular Difference Field (GADF) algorithms, which effectively maps the multidimensional time series constructed for each pixel onto two-dimensional image features, enabling precise extraction and recognition in the backend network algorithms and improving the classification accuracy of land cover types. Secondly, to capture the relationships between features at different scales, this paper proposes a SpectralFormer network architecture using the Context and Structure Encoding (CASE) module to effectively learn dependencies between channels. This architecture enhances important features and suppresses unimportant ones, thereby addressing the semantic gap and improving the recognition capability of land cover features. Finally, the final prediction results are determined by a voting mechanism from the Random Forest algorithm, which synthesizes predictions from multiple decision trees to enhance classification stability and accuracy. To better compare the performance of RS-Net, this paper conducted extensive experiments on three benchmark HS datasets obtained from satellite and airborne imagers, comparing various classic neural network models. Surprisingly, the RS-Net algorithm achieves high performance and efficiency, offering a new and effective tool for land cover classification.

Laplace approximation of J factors for rigid base and rigid basepair models of DNA cyclization

We apply the Laplace approximation to a mathematical formulation of DNA cyclization J factors, leading to a formula that involves energies of local minima of the DNA energy, factors coming from the Hessian of the energy near each minimum, and geometric factors arising from the orientational portion of J. The approximation is derived in a quite general setting that encompasses both rigid base and rigid basepair models common in the literature. The approximation is applied to several families of 200–400 bp DNA, some relatively straight (fragments of λ-phage) and others quite bent (constructs that include up to 10 A tracts). The accuracy of the approximation is assessed by comparing with (more time-consuming) Monte Carlo computations: Laplace is within 20% of Monte Carlo for most 200 bp molecules and undershoots Monte Carlo by about 30% for 300 bp and 50% for 400 bp. We explore length and sequence dependence, both for our overall approximation of J and for its energy and entropic components, and make comparisons to a different approximation of J proposed in the literature.

MSCL-Attention: A Multi-Scale Convolutional Long Short-Term Memory (LSTM) Attention Network for Predicting CO2 Emissions from Vehicles

The transportation industry is one of the major sources of energy consumption and CO2 emissions, and these emissions have been increasing year by year. Vehicle exhaust emissions have had serious impacts on air quality and global climate change, with CO2 emissions being one of the primary causes of global warming. In order to accurately predict the CO2 emission level of automobiles, an MSCL-Attention model based on a multi-scale convolutional neural network, long short-term memory network and multi-head self-attention mechanism is proposed in this study. By combining multi-scale feature extraction, temporal sequence dependency processing, and the self-attention mechanism, the model enhances the prediction accuracy and robustness. In our experiments, the MSCL-Attention model is benchmarked against the latest state-of-the-art models in the field. The results indicate that the MSCL-Attention model demonstrates superior performance in the task of CO2 emission prediction, surpassing the leading models currently available. This study provides a new method for predicting vehicle exhaust emissions, with significant application prospects, and is expected to contribute to reducing global vehicle emissions, improving air quality, and addressing climate change.