-
IEEE Transactions on Pattern Analysis... Jul 2024Dynamic 3D point cloud sequences serve as one of the most common and practical representation modalities of dynamic real-world environments. However, their unstructured...
Dynamic 3D point cloud sequences serve as one of the most common and practical representation modalities of dynamic real-world environments. However, their unstructured nature in both spatial and temporal domains poses significant challenges to effective and efficient processing. Existing deep point cloud sequence modeling approaches imitate the mature 2D video learning mechanisms by developing complex spatio-temporal point neighbor grouping and feature aggregation schemes, often resulting in methods lacking effectiveness, efficiency, and expressive power. In this paper, we propose a novel generic representation called Structured Point Cloud Videos (SPCVs). Intuitively, by leveraging the fact that 3D geometric shapes are essentially 2D manifolds, SPCV re-organizes a point cloud sequence as a 2D video with spatial smoothness and temporal consistency, where the pixel values correspond to the 3D coordinates of points. The structured nature of our SPCV representation allows for the seamless adaptation of well-established 2D image/video techniques, enabling efficient and effective processing and analysis of 3D point cloud sequences. To achieve such re-organization, we design a self-supervised learning pipeline that is geometrically regularized and driven by self-reconstructive and deformation field learning objectives. Additionally, we construct SPCV-based frameworks for both low-level and high-level 3D point cloud sequence processing and analysis tasks, including action recognition, temporal interpolation, and compression. Extensive experiments demonstrate the versatility and superiority of the proposed SPCV, which has the potential to offer new possibilities for deep learning on unstructured 3D point cloud sequences. Code will be released at https://github.com/ZENGYIMING-EAMON/SPCV.
PubMed: 38954587
DOI: 10.1109/TPAMI.2024.3421359 -
IEEE/ACM Transactions on Computational... Jul 2024Biomedical evidence has demonstrated the relevance of microRNA (miRNA) dysregulation in complex human diseases, and determining the relationship between miRNAs and...
Biomedical evidence has demonstrated the relevance of microRNA (miRNA) dysregulation in complex human diseases, and determining the relationship between miRNAs and diseases can aid in the early detection and prevention of diseases. Traditional biological experimental methods have the disadvantages of high cost and low efficiency, which are well compensated by computational methods. However, many computational methods have the challenge of excessively focusing on the neighbor relationship, ignoring the structural information of the graph, and belittling the redundant information of the graph structure. This study proposed a computational model based on a graph-masking autoencoder named MGAEMDA. MGAEMDA is an asymmetric framework in which the encoder maps partially observed graphs into latent representations. The decoder reconstructs the masked structural information based on the edge and node levels and combines it with linear matrices to obtain the result. The empirical results on the two datasets reveal that the MGAEMDA model performs better than its counterparts. We also demonstrated the predictive performance of MGAEMDA using a case study of four diseases, and all the top 30 predicted miRNAs were validated in the database, providing further evidence of the excellent performance of the model.
PubMed: 38954583
DOI: 10.1109/TCBB.2024.3421924 -
IEEE Journal of Biomedical and Health... Jul 2024Deep learning methods have recently achieved remarkable performance in vessel segmentation applications, yet require numerous labor-intensive labeled data. To alleviate...
Deep learning methods have recently achieved remarkable performance in vessel segmentation applications, yet require numerous labor-intensive labeled data. To alleviate the requirement of manual annotation, transfer learning methods can potentially be used to acquire the related knowledge of tubular structures from public large-scale labeled vessel datasets for target vessel segmentation in other anatomic sites of the human body. However, the cross-anatomy domain shift is a challenging task due to the formidable discrepancy among various vessel structures in different anatomies, resulting in the limited performance of transfer learning. Therefore, we propose a cross-anatomy transfer learning framework for 3D vessel segmentation, which first generates a pre-trained model on a public hepatic vessel dataset and then adaptively fine-tunes our target segmentation network initialized from the model for segmentation of other anatomic vessels. In the framework, the adaptive fine-tuning strategy is presented to dynamically decide on the frozen or fine-tuned filters of the target network for each input sample with a proxy network. Moreover, we develop a Gaussian-based signed distance map that explicitly encodes vessel-specific shape context. The prediction of the map is added as an auxiliary task in the segmentation network to capture geometry-aware knowledge in the fine-tuning. We demonstrate the effectiveness of our method through extensive experiments on two small-scale datasets of coronary artery and brain vessel. The results indicate the proposed method effectively overcomes the discrepancy of cross-anatomy domain shift to achieve accurate vessel segmentation for these two datasets.
PubMed: 38954568
DOI: 10.1109/JBHI.2024.3422177 -
IEEE Journal of Biomedical and Health... Jul 2024Estimating blood pressure (BP) values from physiological signals (e.g., photoplethysmogram (PPG)) using deep learning models has recently received increased attention,...
Estimating blood pressure (BP) values from physiological signals (e.g., photoplethysmogram (PPG)) using deep learning models has recently received increased attention, yet challenges remain in terms of models' generalizability. Here, we propose taking a new approach by framing the problem as tracking the "changes" in BP over an interval, rather than directly estimating its value. Indeed, continuous monitoring of acute changes in BP holds promising implications for clinical applications (e.g., hypertensive emergencies). As a solution, we first present a self-contrastive masking (SCM) model, designed to perform pair-wise temporal comparisons within the input signal. We then leverage the proposed SCM model to introduce ΔBPNet, a model trained to detect elevations/drops greater than a given threshold in the systolic blood pressure (SBP) over an interval, from PPG. Using data from PulseDB, 1) we evaluate the performance of ΔBP-Net on previously unseen subjects, 2) we test ΔBP-Net's ability to generalize across domains by training and testing on different datasets, and 3) we compare the performance of ΔBP-Net with existing PPG-based BP-estimation models in detecting over-threshold SBP changes. Formulating the problem as a binary classification task (i.e., over-threshold SBP elevation/ drop or not), ΔBP-Net achieves 75.97%/73.19% accuracy on data from subjects unseen during training. Additionally, the proposed ΔBP-Net outperforms ΔSBP estimations derived from existing PPG-based BP-estimation methods. Overall, by shifting the focus from estimating the value of SBP to detecting overthreshold "changes" in SBP, this work introduces a new potential for using PPG in clinical BP monitoring, and takes a step forward in addressing the challenges related to the generalizability of PPG-based BP-estimation models.
PubMed: 38954566
DOI: 10.1109/JBHI.2024.3422023 -
IEEE Journal of Biomedical and Health... Jul 2024Synergistic drug combination prediction tasks based on the computational models have been widely studied and applied in the cancer field. However, most of models only...
Synergistic drug combination prediction tasks based on the computational models have been widely studied and applied in the cancer field. However, most of models only consider the interactions between drug pairs and specific cell lines, without taking into account the multiple biological relationships of drug-drug and cell line-cell line that also largely affect synergistic mechanisms. To this end, here we propose a multi-modal deep learning framework, termed MDNNSyn, which adequately applies multi-source information and trains multi-modal features to infer potential synergistic drug combinations. MDNNSyn extracts topology modality features by implementing the multi-layer hypergraph neural network on drug synergy hypergraph and constructs semantic modality features through similarity strategy. A multi-modal fusion network layer with gated neural network is then employed for synergy score prediction. MDNNSyn is compared to five classic and state-of-the-art prediction methods on DrugCombDB and Oncology-Screen datasets. The model achieves area under the curve (AUC) scores of 0.8682 and 0.9013 on two datasets, an improvement of 3.70% and 2.71% over the second-best model. Case study indicates that MDNNSyn is capable of detecting potential synergistic drug combinations.
PubMed: 38954565
DOI: 10.1109/JBHI.2024.3421916 -
Nutrition Reviews Jul 2024Time-restricted feeding (TRF) is a lifestyle intervention that aims to maintain a consistent daily cycle of feeding and fasting to support robust circadian rhythms....
Time-restricted feeding (TRF) is a lifestyle intervention that aims to maintain a consistent daily cycle of feeding and fasting to support robust circadian rhythms. Recently, it has gained scientific, medical, and public attention due to its potential to enhance body composition, extend lifespan, and improve overall health, as well as induce autophagy and alleviate symptoms of diseases like cardiovascular diseases, type 2 diabetes, neurodegenerative diseases, cancer, and ischemic injury. However, there is still considerable debate on the primary factors that contribute to the health benefits of TRF. Despite not imposing strict limitations on calorie intake, TRF consistently led to reductions in calorie intake. Therefore, while some studies suggest that the health benefits of TRF are primarily due to caloric restriction (CR), others argue that the key advantages of TRF arise not only from CR but also from factors like the duration of fasting, the timing of the feeding period, and alignment with circadian rhythms. To elucidate the roles and mechanisms of TRF beyond CR, this review incorporates TRF studies that did not use CR, as well as TRF studies with equivalent energy intake to CR, which addresses the previous lack of comprehensive research on TRF without CR and provides a framework for future research directions.
PubMed: 38954563
DOI: 10.1093/nutrit/nuae074 -
Journal of Chemical Theory and... Jul 2024Our ability to calculate rate constants of biochemical processes using molecular dynamics simulations is severely limited by the fact that the time scales for reactions,...
Our ability to calculate rate constants of biochemical processes using molecular dynamics simulations is severely limited by the fact that the time scales for reactions, or changes in conformational state, scale exponentially with the relevant free-energy barrier heights. In this work, we improve upon a recently proposed rate estimator that allows us to predict transition times with molecular dynamics simulations biased to rapidly explore one or several collective variables (CVs). This approach relies on the idea that not all bias goes into promoting transitions, and along with the rate, it estimates a concomitant scale factor for the bias termed the "CV biasing efficiency" γ. First, we demonstrate mathematically that our new formulation allows us to derive the commonly used Infrequent Metadynamics (iMetaD) estimator when using a perfect CV, where γ = 1. After testing it on a model potential, we then study the unfolding behavior of a previously well characterized coarse-grained protein, which is sufficiently complex that we can choose many different CVs to bias, but which is sufficiently simple that we are able to compute the unbiased rate directly. For this system, we demonstrate that predictions from our new Exponential Average Time-Dependent Rate (EATR) estimator converge to the true rate constant more rapidly as a function of bias deposition time than does the previous iMetaD approach, even for bias deposition times that are short. We also show that the γ parameter can serve as a good metric for assessing the quality of the biasing coordinate. We demonstrate that these results hold when applying the methods to an atomistic protein folding example. Finally, we demonstrate that our approach works when combining multiple less-than-optimal bias coordinates, and adapt our method to the related "OPES flooding" approach. Overall, our time-dependent rate approach offers a powerful framework for predicting rate constants from biased simulations.
PubMed: 38954555
DOI: 10.1021/acs.jctc.4c00425 -
Proceedings of the National Academy of... Jul 2024Protein folding and evolution are intimately linked phenomena. Here, we revisit the concept of exons as potential protein folding modules across a set of 38 abundant and...
Protein folding and evolution are intimately linked phenomena. Here, we revisit the concept of exons as potential protein folding modules across a set of 38 abundant and conserved protein families. Taking advantage of genomic exon-intron organization and extensive protein sequence data, we explore exon boundary conservation and assess the foldon-like behavior of exons using energy landscape theoretic measurements. We found deviations in the exon size distribution from exponential decay indicating selection in evolution. We show that when taken together there is a pronounced tendency to independent foldability for segments corresponding to the more conserved exons, supporting the idea of exon-foldon correspondence. While 45% of the families follow this general trend when analyzed individually, there are some families for which other stronger functional determinants, such as preserving frustrated active sites, may be acting. We further develop a systematic partitioning of protein domains using exon boundary hotspots, showing that minimal common exons correspond with uninterrupted alpha and/or beta elements for the majority of the families but not for all of them.
Topics: Exons; Protein Folding; Humans; Proteins; Evolution, Molecular; Introns
PubMed: 38954548
DOI: 10.1073/pnas.2400151121 -
Proceedings of the National Academy of... Jul 2024Standard deep learning algorithms require differentiating large nonlinear networks, a process that is slow and power-hungry. Electronic (CLLNs) offer potentially fast,...
Standard deep learning algorithms require differentiating large nonlinear networks, a process that is slow and power-hungry. Electronic (CLLNs) offer potentially fast, efficient, and fault-tolerant hardware for analog machine learning, but existing implementations are linear, severely limiting their capabilities. These systems differ significantly from artificial neural networks as well as the brain, so the feasibility and utility of incorporating nonlinear elements have not been explored. Here, we introduce a nonlinear CLLN-an analog electronic network made of self-adjusting nonlinear resistive elements based on transistors. We demonstrate that the system learns tasks unachievable in linear systems, including XOR (exclusive or) and nonlinear regression, without a computer. We find our decentralized system reduces modes of training error in order (mean, slope, curvature), similar to in artificial neural networks. The circuitry is robust to damage, retrainable in seconds, and performs learned tasks in microseconds while dissipating only picojoules of energy across each transistor. This suggests enormous potential for fast, low-power computing in edge systems like sensors, robotic controllers, and medical devices, as well as manufacturability at scale for performing and studying emergent learning.
PubMed: 38954545
DOI: 10.1073/pnas.2319718121 -
Proceedings of the National Academy of... Jul 2024Artificial skins or flexible pressure sensors that mimic human cutaneous mechanoreceptors transduce tactile stimuli to quantitative electrical signals. Conventional...
Artificial skins or flexible pressure sensors that mimic human cutaneous mechanoreceptors transduce tactile stimuli to quantitative electrical signals. Conventional trial-and-error designs for such devices follow a forward structure-to-property routine, which is usually time-consuming and determines one possible solution in one run. Data-driven inverse design can precisely target desired functions while showing far higher productivity, however, it is still absent for flexible pressure sensors because of the difficulties in acquiring a large amount of data. Here, we report a property-to-structure inverse design of flexible pressure sensors, exhibiting a significantly greater efficiency than the conventional routine. We use a reduced-order model that analytically constrains the design scope and an iterative "jumping-selection" method together with a surrogate model that enhances data screening. As an exemplary scenario, hundreds of solutions that overcome the intrinsic signal saturation have been predicted by the inverse method, validating for a variety of material systems. The success in property design on multiple indicators demonstrates that the proposed inverse design is an efficient and powerful tool to target multifarious applications of flexible pressure sensors, which can potentially advance the fields of intelligent robots, advanced healthcare, and human-machine interfaces.
PubMed: 38954542
DOI: 10.1073/pnas.2320222121