VEHiCLE: a Variationally Encoded Hi-C Loss Enhancement algorithm for improving and generating Hi-C data.
Ontology highlight
ABSTRACT:
SUBMITTER: Highsmith M
PROVIDER: S-EPMC8065109 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Nowadays, many smart-phones and vehicles are equipped with Global Position System (GPS) for tracking and navigation purposes, providing an opportunity to derive highly representative local vehicular flow and estimate vehicular emissions information. Here, we report and discuss methods used to handle large volumes of such activity data, namely 124 million GPS recordings from the web page Maplink.com.br, extract high spatial resolution vehicular flow information for a vast area in South-east Brazil, and correct for bias using traffic counts observations for the same area. The method consists in filter speed and accelerations, assign buffers to the road network, aggregate speed by street, fill missing number of lanes, generate traffic flow. Methods presented here were used to inform traffic-related air quality modelling and used as part of local air pollution management activities but are also amenable to any work that would be enhanced by more locally representative or time-resolved inputs for traffic flow, e.g. traffic network management, and demand modelling. •124 million GPS observations from electronic devices were used to generate traffic flow.•Spatial bias was investigated and accounted for using independent local traffic count data.•Traffic count rescaled GPS traffic flow provide a robust description of spatial and quantitative traffic patterns.
| S-EPMC6812335 | biostudies-literature
Project description:Collective variable-based enhanced sampling methods are routinely used on systems with metastable states, where high free energy barriers impede the proper sampling of the free energy landscapes when using conventional molecular dynamics simulations. One such method is variationally enhanced sampling (VES), which is based on a variational principle where a bias potential in the space of some chosen slow degrees of freedom, or collective variables, is constructed by minimizing a convex functional. In practice, the bias potential is taken as a linear expansion in some basis function set. So far, primarily basis functions delocalized in the collective variable space, like plane waves, Chebyshev, or Legendre polynomials, have been used. However, there has not been an extensive study of how the convergence behavior is affected by the choice of the basis functions. In particular, it remains an open question if localized basis functions might perform better. In this work, we implement, tune, and validate Daubechies wavelets as basis functions for VES. The wavelets construct orthogonal and localized bases that exhibit an attractive multiresolution property. We evaluate the performance of wavelet and other basis functions on various systems, going from model potentials to the calcium carbonate association process in water. We observe that wavelets exhibit excellent performance and much more robust convergence behavior than all other basis functions, as well as better performance than metadynamics. In particular, using wavelet bases yields far smaller fluctuations of the bias potential within individual runs and smaller differences between independent runs. Based on our overall results, we can recommend wavelets as basis functions for VES.
| S-EPMC9281396 | biostudies-literature
Project description:Lack of attention to missing data in research may result in biased results, loss of power and reduced generalizability. Registering reasons for missing values at the time of data collection, or-in the case of sharing existing data-before making data available to other teams, can save time and efforts, improve scientific value and help to prevent erroneous assumptions and biased results. To ensure that encoding of missing data is sufficient to understand the reason why data are missing, it should ideally be context-free. Therefore, 11 context-free codes of missing data were carefully designed based on three completed randomized controlled clinical trials and tested in a new randomized controlled clinical trial by an international team consisting of clinical researchers and epidemiologists with extended experience in designing and conducting trials and an Information System expert. These codes can be divided into missing due to participant and/or participation characteristics (n = 6), missing by design (n = 4), and due to a procedural error (n = 1). Broad implementation of context-free missing data encoding may enhance the possibilities of data sharing and pooling, thus allowing more powerful analyses using existing data.
| S-EPMC5595279 | biostudies-literature
Project description:Estimating truck emissions accurately would benefit atmospheric research and public health protection. Here, we developed a full-sample enumeration approach TrackATruck to bridge low-frequency but full-size vehicles driving big data to high-resolution emission inventories. Based on 19 billion trajectories, we show how big the emission difference could be using different approaches: 99% variation coefficients on regional total (including 31% emissions from non-local trucks), and ± as large as 15 times on individual counties. Even if total amounts are set the same, the emissions on primary cargo routes were underestimated in the former by a multiple of 2-10 using aggregated approaches. Time allocation proxies are generated, indicating the importance of day-to-day estimation because the variation reached 26-fold. Low emission zone policy reduced emissions in the zone, but raised emissions in upwind areas in Beijing's case. Comprehensive measures should be considered, e.g. the demand-side optimization.
| S-EPMC7271216 | biostudies-literature
Project description:MotivationHow chromatin folds in three-dimensional (3D) space is closely related to transcription regulation. As powerful tools to study such 3D chromatin conformation, the recently developed Hi-C technologies enable a genome-wide measurement of pair-wise chromatin interaction. However, methods for the detection of biologically meaningful chromatin interactions, i.e. peak calling, from Hi-C data, are still under development. In our previous work, we have developed a novel hidden Markov random field (HMRF) based Bayesian method, which through explicitly modeling the non-negligible spatial dependency among adjacent pairs of loci manifesting in high resolution Hi-C data, achieves substantially improved robustness and enhanced statistical power in peak calling. Superior to peak callers that ignore spatial dependency both methodologically and in performance, our previous Bayesian framework suffers from heavy computational costs due to intensive computation incurred by modeling the correlated peak status of neighboring loci pairs and the inference of hidden dependency structure.ResultsIn this work, we have developed FastHiC, a novel approach based on simulated field approximation, which approximates the joint distribution of the hidden peak status by a set of independent random variables, leading to more tractable computation. Performance comparisons in real data analysis showed that FastHiC not only speeds up our original Bayesian method by more than five times, bus also achieves higher peak calling accuracy.Availability and implementationFastHiC is freely accessible at:http://www.unc.edu/∼yunmli/FastHiC/ CONTACTS: : yunli@med.unc.edu or ming.hu@nyumc.orgSupplementary informationSupplementary data are available at Bioinformatics online.
| S-EPMC5013904 | biostudies-literature
Project description:Vision based vehicle detection is a critical technology that plays an important role in not only vehicle active safety but also road video surveillance application. Traditional shallow model based vehicle detection algorithm still cannot meet the requirement of accurate vehicle detection in these applications. In this work, a novel deep learning based vehicle detection algorithm with 2D deep belief network (2D-DBN) is proposed. In the algorithm, the proposed 2D-DBN architecture uses second-order planes instead of first-order vector as input and uses bilinear projection for retaining discriminative information so as to determine the size of the deep architecture which enhances the success rate of vehicle detection. On-road experimental results demonstrate that the algorithm performs better than state-of-the-art vehicle detection algorithm in testing data sets.
| S-EPMC4052056 | biostudies-other
Project description:Sampling complex free-energy surfaces is one of the main challenges of modern atomistic simulation methods. The presence of kinetic bottlenecks in such surfaces often renders a direct approach useless. A popular strategy is to identify a small number of key collective variables and to introduce a bias potential that is able to favor their fluctuations in order to accelerate sampling. Here, we propose to use machine-learning techniques in conjunction with the recent variationally enhanced sampling method [O. Valsson, M. Parrinello, Phys. Rev. Lett. 113, 090601 (2014)] in order to determine such potential. This is achieved by expressing the bias as a neural network. The parameters are determined in a variational learning scheme aimed at minimizing an appropriate functional. This required the development of a more efficient minimization technique. The expressivity of neural networks allows representing rapidly varying free-energy surfaces, removes boundary effects artifacts, and allows several collective variables to be handled.
| S-EPMC6731643 | biostudies-literature
Project description:For family-based association studies, Horvath et al. proposed an algorithm for the association analysis between haplotypes and arbitrary phenotypes when the phase of the haplotypes is unknown, that is, genotype data is given. Their approach to haplotype analysis maintains the original features of the TDT/FBAT-approach, that is, complete robustness against genetic confounding and misspecification of the phenotype. The algorithm has been implemented in the FBAT and PBAT software package and has been used in numerous substantive manuscripts. Here, we propose a simplification of the original algorithm that maintains the original approach but reduces the computational burden of the approach substantially and gives valuable insights regarding the conditional distribution. With the modified algorithm, the application to whole-genome sequencing (WGS) studies becomes feasible; for example, in sliding window approaches or spatial-clustering approaches. The reduction of the computational burden that our modification provides is especially dramatic when both parental genotypes are missing. For example, for eight variants and 441 nuclear families with mostly offspring-only families, in a WGS study at the APOE locus, the running time decreased from approximately 21 hr for the original algorithm to 0.11 sec after our modification.
| S-EPMC5774664 | biostudies-literature
Project description:High-throughput assays for measuring the three-dimensional (3D) configuration of DNA have provided unprecedented insights into the relationship between DNA 3D configuration and function. Data interpretation from assays such as ChIA-PET and Hi-C is challenging because the data is large and cannot be easily rendered using standard genome browsers. An effective Hi-C visualization tool must provide several visualization modes and be capable of viewing the data in conjunction with existing, complementary data. We review five software tools that do not require programming expertise. We summarize their complementary functionalities, and highlight which tool is best equipped for specific tasks.
| S-EPMC5290626 | biostudies-literature
Project description:Deep Convolutional Neuronal Networks (DCNNs) are showing remarkable performance on many computer vision tasks. Due to their large parameter space, they require many labeled samples when trained in a supervised setting. The costs of annotating data manually can render the use of DCNNs infeasible. We present a novel framework called RenderGAN that can generate large amounts of realistic, labeled images by combining a 3D model and the Generative Adversarial Network framework. In our approach, image augmentations (e.g., lighting, background, and detail) are learned from unlabeled data such that the generated images are strikingly realistic while preserving the labels known from the 3D model. We apply the RenderGAN framework to generate images of barcode-like markers that are attached to honeybees. Training a DCNN on data generated by the RenderGAN yields considerably better performance than training it on various baselines.
| S-EPMC7805882 | biostudies-literature