Find research datasets worth reusing

DeepLabCut network trained to track mouse 'back' body parts during rotarod running (back-view)

DeepLabCut (https://github.com/DeepLabCut/) (Mathis et al., 2018; Nath et al., 2019) was used for tracking body parts of mice in an open field arena or in the rotarod. DeepLabCut 2.1.8.2 (local version on Windows with CPU, using the GUI) and 2.1.10.2 (google colab to train the network) were used using default parameters and the pretrained resnet50 network with imgaug augmentation. Frames were extracted with the k-means method and outlier frames with the jump method. Rotarod, back camera: 20 images from 9 videos (10 fps) were extracted for a total of 180 labeled pictures. 5 body parts (2 paws, 2 ankles, tail base), 4 corners of the rotarod, 2 points on the rotarod wheels and 4 points in a flashing LED (indicating timestamps) were manually labeled. A neural network was trained using these images for 80K iterations. 20 images from 14 videos with different recording conditions were labeled. The network was trained to 200K iterations (from scratch) (train error: 3.00, test error: 3.75). Relevant videos were analyzed at each step, for a total of 152 videos. Used to analyze videos for a publication (Labouesse et al., Nature Communications 2023)

Supplementary Material of : Large-Scale 3D Image Segmentation Using Scattering Networks

The reader will find here the supplementary material associated with the manuscript "Large-Scale 3D Image Segmentation Using Scattering Networks" submitted to IEEE Transaction of Pattern Analysis and Machine Intelligence (TPAMI), 2022.

A Deep Neural Network Based SMAP Soil Moisture Product

It is demonstrated that while satellite soil moisture (SM) retrievals often have minimum biases, reanalysis data can capture more temporal variability of SM, especially for non-cropland areas -- when validated against in situ measurements. Accordingly, this paper presents a deep neural network (DNN) that utilizes the merits of a suite of existing satellite and reanalysis products to produce a new SM product with minimum (maximum) bias (correlation) -- using NASA’s Soil Moisture Active Passive (SMAP) data and ERA5 reanalysis. The benchmark of the network is a bias-adjusted SM with maximum correlation with in situ data over each land-cover type. The mean of the benchmark data is adjusted to the product that exhibits a minimum bias over each land-cover type. Consistent with the laws of L-band microwave propagation in soil and canopy, the input variables of DNN include polarized SMAP brightness temperatures, incidence angle, vegetation scattering albedo, surface roughness parameter, surface water fraction, effective soil temperatures, bulk density, clay fraction, and vegetation optical depth from the normalized difference vegetation index (NDVI) climatology. The DNN is trained and validated using two years (04/2015--03/2017) of global data and deployed for assessment of its performance from 04/2017 to 03/2021. The testing results against in situ measurements demonstrate that the DNN outputs typically exhibit improved error quality metrics over most land-cover types and climate regimes and can properly capture SM temporal dynamics, beyond each SMAP product across regional to continental scales.

MCPNet : A parallel maximum capacity-based genome-scale gene network construction framework

This deposit contains the gene expression profile datasets used for the paper titled "MCPNet : A parallel maximum capacity-based genome-scale gene network construction framework".  There are three sets of data: <ul> <li>Simulated yeast data from NetBenchmark, with random noise injected, as well as the ground truth network matrix.  In "SimulatedYeast.zip".</li> <li>Real Yeast dataset and the ground truth network as an adjacency list file.  in "yeast_data.exp" and "yeast_gs1_list_filtered.tsv".  Data acquired from "Castro DM, de Veaux NR, Miraldi ER, Bonneau R (2019) Multi-study inference of regulatory networks for more accurate models of gene regulation. PLoS Comput Biol 15(1): e1006591. https://doi.org/10.1371/journal.pcbi.1006591", <a href="https://github.com/simonsfoundation/multitask_inferelator/tree/AMuSR">https://github.com/simonsfoundation/multitask_inferelator/tree/AMuSR</a>.</li> <li>Real Arabidopsis athaliana datasets for 5 tissues and 1 environmental challenge.  <ul> <li>athaliana_gs_probes.tsv : ground truth as an adjacency list</li> <li>microarray gene expression profiles for "development", "leaf", "seed", "flower", "seedling1week", "hormone-aba-iaa-ga-br".</li> <li>Aathaliana.Datasets-CEL-File-URLs.xlsx: list of SRA accession numbers for the A. athaliana datasets</li> </ul> </li> </ul>

opencc-zeroApr 2022View details →

R_JAGS code for estimation and analysis of species-area-relationship (SAR) parameters from NEON (National Ecological Observatory Network) data on plant surveys

Invasive species science is heavily geared toward the invasive agent. However, management to protect native species also requires a proactive approach focused on understanding the features affecting community vulnerability to invasion impacts. Vulnerability is likely the result of factors acting across spatial scales, from local to regional, and it is the combined effects of these factors that will determine the magnitude of vulnerability. We introduce an analytical framework that quantifies the scale-dependent impact of biological invasions from the shape of the native species-area-relationship (SAR). We leverage newly available, biogeographically extensive vegetation data from the US National Ecological Observatory Network to assess plant community vulnerability to invasion impact as a function of factors acting across scales. We analyzed more than 1000 SARs widely distributed across the USA along environmental gradients and under different levels of invasion. Results show that a decrease in native richness is consistently associated with invasive species cover, but it is only at relatively high levels of invasion that native richness is compromised. After accounting for variation in baseline ecosystem diversity, net primary productivity, and human modification, ecoregions that are colder and wetter seem to be most vulnerable to losses of native plant species at the local level, while warmer and wetter areas seem most susceptible at the landscape level. We also document how the combined effects of cross-scale factors result in a heterogenous spatial pattern of vulnerability. This pattern cannot be predicted by analyses at any single scale, underscoring the importance of accounting for factors acting across scales. Simultaneously assessing differences in vulnerability between distinct plant communities at local, landscape and regional scales provided outputs that can be used to inform policy and management aimed at reducing vulnerability to the impact of plant invasions.

Designing of Fiber Bragg Gratings for Long-distance Optical Fiber Sensing Networks

Research data of Modelling and Simulation in Engineering journal article “Designing of Fiber Bragg Gratings for Long-distance Optical Fiber Sensing Networks”. Most optical sensors on the market are optical fiber Bragg grating (FBG) sensors with low reflectivity (typically 7-40%) and low side-lobe suppression (SLS) ratio (typically SLS <15dB), which prevents these sensors from being effectively used for long-distance remote monitoring and sensor network solutions. This research is based on designing the optimal grating structure of FBG sensors and estimating their optimal apodization parameters necessary for sensor networks and long-distance monitoring solutions. Gaussian, sine and raised sine apodizations are studied to achieve the main requirements, which are - maximally high reflectivity (at least 90%) and side-lobe suppression (at least 20 dB), as well as maximally narrow bandwidth (FWHM<0.2 nm), FBGs with uniform (without apodization). Results gathered in this research propose high-efficiency FBG grating apodizations, which can be further physically realized for optical sensor networks and long-distance (at least 40 km) monitoring solutions.

Stable species and interactions in plant-pollinator networks deviate from core position in fragmented habitats

Species and their interactions are more dynamic over time and space in fragmented habitats than in continuous habitats. In fragmented habitats, the low nestedness of mutualistic networks may be related to the position change of stable (high persistence over time/space) species and interactions in the networks. Previous studies have shown that stable species and interactions tend to be in the core position of mutualistic networks. However, in fragmented habitats, it remains unknown whether stable species or interactions still tend to be in the core position. To address this gap, here we evaluated the correlation between the position of proximity to the network core and the temporal/spatial stability of species and interactions, using the observation of 42 plant-pollinator networks conducted in a fragmented island landscape over 3 years. We showed that temporally/spatially stable species and interactions deviated from the network core to varying degrees. Temporally stable plants were most likely to deviate from the network core, followed by pollinators and interactions, while only spatially stable pollinators tend to deviate from the network core. When unstable species (present in few time/space points, typically specialists) and interactions occupy the network core, they cannot interact with most species in the network as generalists do, resulting in the decrease of network nestedness. Therefore, from the perspective of position and stability, stable species and interactions deviate from the network core in fragmented habitats, which is an important reason for the decrease of nestedness in mutualistic networks. Our study suggests that protecting plants that occupy the core in large plant-pollinator networks is essential for maintaining the network persistence in fragmented habitats.

A tempοral Deep Convolutional Neural Network model on Sentinel-1 Image Time Series for pixel-wise Flood Classification (dataset)

This is a dataset which has been designed to be used for flood time series classification. Each time series is annotated as flood or no-flood and represents a pixel-wise time series derived from stack of Sentinel-1 IW GRD images that have been pre-processed according to <a href="http://doi.org/10.5281/zenodo.6510223">https://doi.org/10.5281/zenodo.6510223</a>.

Nest choice in arboreal ants is an emergent consequence of network creation under spatial constraints

Biological transportation networks must balance competing functional priorities. The self-organizing mechanisms used to generate such networks have inspired scalable algorithms to construct and maintain low-cost and efficient human-designed transport networks. The pheromone-based trail networks of ants have been especially valuable in this regard. Here, we use turtle ants as our focal system: In contrast to the ant species usually used as models for self-organized networks, these ants live in a spatially constrained arboreal environment where both nesting options and connecting pathways are limited. Thus, they must solve a distinct set of challenges which resemble those faced by human transport engineers constrained by existing infrastructure. Here, we ask how a turtle ant colony's choice of which nests to include in a network may be influenced by their potential to create connections to other nests. In laboratory experiments with Cephalotes varians and Cephalotes texanus, we show that nest choice is influenced by spatial constraints, but in unexpected ways. Under one spatial configuration, colonies preferentially occupied more connected nest sites; however, under another spatial configuration, this preference disappeared. Comparing the results of these experiments to an agent-based model, we demonstrate that this apparently idiosyncratic relationship between nest connectivity and nest choice can emerge without nest preferences via a combination of self-reinforcing random movement along constrained pathways and density-dependent aggregation at nests. While this mechanism does not consistently lead to the de-novo construction of low-cost, efficient transport networks, it may be an effective way to expand a network, when coupled with processes of pruning and restructuring.

Parallel generation of extensive vascular networks with application to an archetypal human kidney model

Given the relevance of the inextricable coupling between microcirculation and physiology, and the relation to organ function and disease progression, the construction of synthetic vascular networks for mathematical modelling and computer simulation is becoming an increasingly broad field of research. Building vascular networks that mimic in-vivo morphometry is feasible through algorithms such as constrained constructive optimisation (CCO) and variations. Nevertheless, these methods are limited by the maximum number of vessels to be generated due to the whole network update required at each vessel addition. In this work, we propose a CCO-based approach endowed with a domain decomposition strategy to concurrently create vascular networks. The performance of this approach is evaluated by analysing the agreement with the sequentially generated networks and studying the scalability when building vascular networks up to 200,000 vascular segments. Finally, we apply our method to vascularise a highly complex geometry corresponding to the cortex of a prototypical human kidney. The technique presented in this work enables the automatic generation of extensive vascular networks, removing the limitation from previous works. Thus, we can extent vascular networks (e.g., obtained from medical images) to pre-arteriolar level, yielding patient-specific whole-organ vascular models with an unprecedented level of detail.

SUBATOMIC analysis of Homo sapiens integrated multi-edge networks

We applied SUBATOMIC (https://github.com/CBIGR/SUBATOMIC/) to analyze a composite H. sapiens network containing transcription factor-target gene, miRNA-target gene, protein-protein, homologous and co-functional interactions from three different databases. We derived and annotated 5586 modules with diverse topological, regulatory and functional properties.

opencc-by-4.0Mar 2022View details →

Graphic Novel Character Networks and Statistics

Description. This dataset contains the character networks extracted from the graphic novel Thorgal, as well as the statistics and plots produced when analyzing these networks. Source code. The source code used to produce these files is available on GitHub: <a href="https://github.com/CompNet/NaNet">https://github.com/CompNet/NaNet</a> Citation. If you use these data, please cite the following article: <ul> <li>V. Labatut, “Complex Network Analysis of a Graphic Novel: The Case of the Bande Dessinée Thorgal,” Advances in Complex Systems, p. 22400033, 2022. ⟨<a href="https://hal.archives-ouvertes.fr/hal-03694768">hal-03694768</a>⟩ - DOI: <a href="http://doi.org/10.1142/S0219525922400033">10.1142/S0219525922400033</a></li> </ul> <code>@Article{Labatut2022,</code> <code>  author    = {Labatut, Vincent},</code> <code>  title     = {Complex Network Analysis of a Graphic Novel: The Case of the Bande Dessinée {T}horgal},</code> <code>  journal   = {Advances in Complex Systems},</code> <code>  year      = {2022},</code> <code>  volume    = {25},</code> <code>  number    = {5\&6},</code> <code>  pages     = {2240003},</code> <code>  doi       = {10.1142/S0219525922400033},</code> <code>}</code>

Decomposition, topology, properties, and graphs of woody crown networks of 15 tree species of Cerrado vegetation

Data of decomposition, topology, properties, and the corresponding graphs of 15 adult tree species of Cerrado vegetation, sensu stricto physiognomy. The woody crown networks (WCN) representations in a bidimensional space were obtained by drawing followed the methodology described by Prado et al. (2020, Prado, C.H.B.A., Trovão, D.M.B.M., Souza, J.P., 2020. A network model for determining the woody crown's decomposition, topology, and properties. Journal of Theoretical Biology, v. 499, p. 110318. https://doi.org/<a href="https://www.x-mol.com/paperRedirect/1258515479077781504">10.1016/j.jtbi.2020.110318</a>.). The branching regions were the nodes, and the woody crown segments connecting the nodes or merely emerging from them were the connectors. Those trees grew under natural conditions in a most common (sensu stricto) physiognomy of Cerrado vegetation, in a reservoir of 86 ha, located at 850 m above sea level in São Carlos city, São Paulo state, Brazil, at 21°58'- 22°00'S and 47°51'-47°52'W. Following the Köppen climatic classification, this region is between Aw and Cwa, a tropical climate with dry winter and wet summer. The rainy season occurs between October-March, and the dry season between April and September.

Temporal and spatial changes in benthic invertebrate trophic networks along a taxonomic richness gradient

Species interactions underlie most ecosystem functions and are important for understanding ecosystem changes. Representing one type of species interaction, trophic networks were constructed from biodiversity monitoring data and known trophic links to assess how ecosystems have changed over time. The Baltic Sea is subject to many anthropogenic pressures, and low species diversity makes it an ideal candidate for determining how pressures change food webs. In this study, we used benthic monitoring data from 20 years (1980-1989 and 2010-2019) from the Swedish coast of the Baltic Sea and Skagerrak to investigate changes in benthic invertebrate trophic interactions. We constructed food webs and calculated fundamental food web metrics evaluating network horizontal and vertical diversity, as well as stability that were compared over space and time. Our results show that the west coast of Sweden (Skagerrak) suffered a reduction in benthic invertebrate biodiversity by 32 % between the 1980's and 2010's, and that the number of links, generality of predators, and vulnerability of prey, have been significantly reduced. The other basins (Bothnian Sea, Baltic Proper and Bornholm Basin) do not show any significant changes in species richness or consistent significant trends in any food web metrics investigated, demonstrating resilience at a lower species diversity. The decreased complexity of the Skagerrak food webs indicates vulnerability to further perturbations and pressures should be limited as much as possible to ensure continued ecosystem functions.

Sequence Similarity Network (SSN) and Genome Neighbourhood Network (GNN) for Mycobacterium Cytochrome P450 enzymes

This dataset was generated in the context of the Horizon 2020 MSCA IF action deCrYPtion (Grant 839116). The aim of this project is to use comparative genomics in order to propose and then test the function of uncharacterised Cytochrome P450 enzymes that are present among Mycobacterium species. More information about this project can be found at: https://cordis.europa.eu/project/id/839116. This dataset contains: - The FASTA sequences files obtained from the UniProt database, for members of the PF00067 protein family (CYP). - A set of reference FASTA sequences, matching the supplementary material from the following publication: Parvez, M. et al. (2016) ‘Molecular evolutionary dynamics of cytochrome P450 monooxygenases across kingdoms: Special focus on mycobacterial P450s’, Scientific Reports, 6(1), p. 33099. doi:<a href="https://doi.org/10.1038/srep33099">10.1038/srep33099</a>. - A combined FASTA files of both previously described, that was used for the generation of SSNs - A PNG image produced from the analysis of the Sequence Similarity Networks generated at AST78 (corresponding to 40% identity, defining CYP families) - A PNG image produced from the analysis of the Sequence Similarity Networks generated at AST141 (corresponding to 55% identity, defining CYP subfamilies) - A Cytoscape session for the Sequence Similarity Networks from the combined FASTA file generated using the Enzyme Function Initiative web tools (https://efi.igb.illinois.edu), at AST78 - A Cytoscape session containing Sequence Similarity Networks and Genome Neighborhood Network from the combined FASTA file  generated using the Enzyme Function Initiative web tools (https://efi.igb.illinois.edu), at AST141

Logs and QC information for the northern Borneo Orogeny Seismic Survey seismic network

Instrument log files (mass positions + GPS offset/syncs and system information) for the northern Borneo Orogeny Seismic Survey (nBOSS) seismic network, which operated 2018–2020. FDSN network code: YC (<a href="https://doi.org/10.7914/SN/YC_2018">https://doi.org/10.7914/SN/YC_2018</a>).

opencc-by-4.0Jun 2022View details →

Loon Network Data

Loon sought to provide data connectivity to under-served regions via 4G LTE cell service hosted on free-floating, high-altitude balloons. To backhaul data, the balloons used point-to-point radio links to connect with each other and to transceivers connected to the network core on the ground (aka. "ground stations"). These links were selected & commanded by a centralized Temporo-Spatial Software Defined Network (TS-SDN) controller running on the ground. More details about the system architecture and operation can be found in the <a href="https://x.company/projects/loon/the-loon-collection/">Loon Library</a> or in "SDN in the Stratosphere: Loon's Aerospace Mesh Network" from the proceedings of SIGCOMM 2022.   The dataset provided here consists of internal state from the TS-SDN and network telemetry gathered from serving commercial traffic and R&D experiments. Loon’s networking use case is unique in several ways: <ul> <li> Balloons and their base stations/eNodeBs were located in the stratosphere and changed position as they were pushed by the wind. </li> <li> The network was comprised of a combination of LTE, WiFi, E-band, Wired technologies </li> <li> The topology of our E-band backhaul meshes spanned thousands of kilometers and changed frequently. </li> <li> eNodeBs entered and exited the cellular network depending on their position and the availability of power and backhaul connectivity. </li> </ul>   Loon collected data using system telemetry from Loon assets (balloons, ground stations, etc) in the field. These logs were extracted from various storage systems and processed for external publication.

opencc-by-4.0Jun 2022View details →

LoRaWAN Network Attacks

This dataset includes a dump of messages seen on a LoRaWAN network collected during validation of the GUARD project. It includes plain traffic from real devices installed on city bus and some attacks artificially generated by replicating or altering messages.

opencc-by-4.0Jun 2022View details →

Fireball events detected in Romania by the FRIPON-MOROI network

In this table we present the fireball events detected by the Fireball Recovery and Inter Planetary Observation Network (FRIPON) (Colas et al., 2020) with the use of the Meteorite Orbits Reconstruction by Optical Imaging (MOROI) network (Nedelcu et al. 2018). The data include events recorded by at least two different fireball cameras installed over the Romanian territory. For each event we provide: the name, the stations that detected it, whether it has noticeable deceleration (Vf/V0 < 0.8, where Vf is the velocity of the meteoroid at the end of the luminous trajectory and V0 is the initial velocity of the meteoroid). The events are named by the time when they were captured as follows: year, month, day followed by hour, minute, second in UT. The events with noticeable deceleration are analysed in (Boaca et al. 2022), https://iopscience.iop.org/article/10.3847/1538-4357/ac8542. If using this data, please reference the above publication.  The online presentation of the paper can be found on the AAS youtube channel: https://www.youtube.com/watch?v=Wx186QB5ZVE