Find research datasets worth reusing

opencc-by-4.0Mar 2024View details →

Hourly values of an advanced human-biometeorological index for diverse populations from 1991 to 2020

The presented human thermal bioclimate dataset was created in the frame of the <a href="https://theheatalarm.wordpress.com/">HEAT-ALARM</a> ("Development of a heat-health warning system in Greece") research project. Initially developed for Greece,  it consists of hourly values of population-weighted mPET (modified physiologically equivalent temperature), simulated by the RayMan Pro model for the period 1991-2020 and for 10 population subsets in 72 regional units and combinations thereof, which are based on the NUTS-3 (Nomenclature of Territorial Units for Statistics-3) classification in Greece, using the Copernicus European Regional Reanalysis (CERRA) at 5.5 km spatial resolution. The dataset also includes the main environmental drivers of mPET (e.g. temperature) at the same spatiotemporal resolution. In the framework of replicating the original dataset, the current version includes population-weighted values of mPET and its environmental drivers for six populations in five districts of Cyprus at the LAU-1 (Local Administrative Units-1) level, covering the period from 1991 to 2020.  The code used to produce the presented data is available at: <a href="https://doi.org/10.5281/zenodo.10793067">https://doi.org/10.5281/zenodo.10793067</a>. It can be used to replicate the dataset not only directly in Greece, but also in any other country included in the CERRA domain after appropriate adjustments, as in the case of Cyprus above. Compared to the previous version of the dataset for Greece, this version includes vapor pressure (VP) instead of relative humidity (see README.txt for more details), as VP is more relevant for human-biometerological and health-related studies.  References Giannaros, C., Agathangelidis, I., Galanaki, E. et al. Hourly values of an advanced human-biometeorological index for diverse populations from 1991 to 2020 in Greece. Sci Data 11, 76 (2024). <a href="https://doi.org/10.1038/s41597-024-02923-y">https://doi.org/10.1038/s41597-024-02923-y</a>

opencc-by-4.0May 2022View details →

"Centenarians have a diverse population of gut bacteriophages that may promote healthy lifespan" - Genomes and annotation

File-dump associated with the manuscript: "Centenarians have a diverse population of gut bacteriophages that may promote healthy lifespan" (Not yet published) MGVs refer to the viral genome database in the publication: https://www.nature.com/articles/s41564-021-00928-6    Following uploaded: File 1: VOG Markers in vOTUs/vMAGs and MGV genomes File 2: Viral Tree Newick file with vOTUs/vMAGs and MGV genomes File 3: All vOTUs/vMAGs genomes File 4: Master table annotation of vOTUs/vMAGs File 5: Centenarian bacterial isolate proviruses

Model Zoo: A Dataset of Diverse Populations of Neural Network Models - CIFAR10

Abstract In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before. Dataset This dataset is part of a larger collection of model zoos and contains the zoos trained on CIFAR10. All zoos with extensive information and code can be found at www.modelzoos.cc. This repository contains two types of files: the raw model zoos as collections of models (file names beginning with "cifar_"), as well as preprocessed model zoos wrapped in a custom pytorch dataset class (filenames beginning with "dataset"). Zoos are trained with small and large CNN models, in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). The index_dict.json files contain information on how to read the vectorized models. For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

Model Zoo: A Dataset of Diverse Populations of Neural Network Models - STL10 - Raw Datasets

Abstract In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before. Dataset This dataset is part of a larger collection of model zoos and contains the zoos trained on the labelled samples from STL10. All zoos with extensive information and code can be found at www.modelzoos.cc. This repository contains the raw model zoos as collections of models (file names beginning with "cifar_"). Zoos are trained with small and large CNN models, in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). Due to the large filesize, the preprocessed datasets are hosted in a separate repository. The index_dict.json files contain information on how to read the vectorized models. For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

Model Zoo: A Dataset of Diverse Populations of Neural Network Models - SVHN

Abstract In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before. Dataset This dataset is part of a larger collection of model zoos and contains the zoos trained on the labelled samples from SVHN. All zoos with extensive information and code can be found at www.modelzoos.cc. This repository contains two types of files: the raw model zoos as collections of models (file names beginning with "svhn_"), as well as preprocessed model zoos wrapped in a custom pytorch dataset class (filenames beginning with "dataset"). Zoos are trained in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). The index_dict.json files contain information on how to read the vectorized models. For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

Model Zoo: A Dataset of Diverse Populations of Neural Network Models - Fashion-MNIST

Abstract In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before. Dataset This dataset is part of a larger collection of model zoos and contains the zoos trained on the labelled samples from Fashion-MNIST. All zoos with extensive information and code can be found at www.modelzoos.cc. This repository contains two types of files: the raw model zoos as collections of models (file names beginning with "fmnist_"), as well as preprocessed model zoos wrapped in a custom pytorch dataset class (filenames beginning with "dataset"). Zoos are trained in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). The index_dict.json files contain information on how to read the vectorized models. For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

Model Zoo: A Dataset of Diverse Populations of Neural Network Models - MNIST

Abstract In the last years, neural networks have evolved from laboratory environments to the state-of-the-art for many real-world problems. Our hypothesis is that neural network models (i.e., their weights and biases) evolve on unique, smooth trajectories in weight space during training. Following, a population of such neural network models (refereed to as “model zoo”) would form topological structures in weight space. We think that the geometry, curvature and smoothness of these structures contain information about the state of training and can be reveal latent properties of individual models. With such zoos, one could investigate novel approaches for (i) model analysis, (ii) discover unknown learning dynamics, (iii) learn rich representations of such populations, or (iv) exploit the model zoos for generative modelling of neural network weights and biases. Unfortunately, the lack of standardized model zoos and available benchmarks significantly increases the friction for further research about populations of neural networks. With this work, we publish a novel dataset of model zoos containing systematically generated and diverse populations of neural network models for further research. In total the proposed model zoo dataset is based on six image datasets, consist of 24 model zoos with varying hyperparameter combinations are generated and includes 47’360 unique neural network models resulting in over 2’415’360 collected model states. Additionally, to the model zoo data we provide an in-depth analysis of the zoos and provide benchmarks for multiple downstream tasks as mentioned before. Dataset This dataset is part of a larger collection of model zoos and contains the zoos trained on the labelled samples from MNIST. All zoos with extensive information and code can be found at www.modelzoos.cc. This repository contains two types of files: the raw model zoos as collections of models (file names beginning with "mnist_"), as well as preprocessed model zoos wrapped in a custom pytorch dataset class (filenames beginning with "dataset"). Zoos are trained in three configurations varying the seed only (seed), varying hyperparameters with fixed seeds (hyp_fix) or varying hyperparameters with random seeds (hyp_rand). The index_dict.json files contain information on how to read the vectorized models. For more information on the zoos and code to access and use the zoos, please see www.modelzoos.cc.

opencc-by-4.0Jul 2022View details →

Data for: Seasonal dynamics of faunal diversity and population ecology in an estuarine seagrass bed

These are the data used in the analyses described in the paper titled "Seasonal dynamics of faunal diversity and population ecology in an estuarine seagrass bed", accepted at Estuaries and Coasts. We acknowledge the tangata whenua for the rohe in which these data were collected, Ngāi Tārewa and Ngāti Īrakehu. We thank the Akaroa Taiāpure for their support of this research. The data included are: Raw count data of taxa for each tow, associated with additional metadata including the date of collection, tow coordinates, and estimated seagrass cover (MonthlyRawSampling_Duvauchelle_2020.csv). This data was put through cleaning steps outlined in the file docs/dataCleaning.Rmd prior to being used in any analyses. The cleaned community composition data (cleanedCommunity.csv), output from <a href="https://github.com/spflanagan/ecology-duvauchelle/blob/main/docs/dataCleaning.Rmd">docs/dataCleaning.Rmd</a> and used in the downstream community and population analyses. The GPS coordinates for the tows (gpsdat.csv). These were extracted from the raw data in the data cleaning process. NZsyngnathids_measurements.csv contains the measurements of the pipefish from images. These data also underwent a cleaning process documented in <a href="https://github.com/spflanagan/ecology-duvauchelle/blob/main/docs/dataCleaning.Rmd">docs/dataCleaning.Rmd</a>. The cleaned pipefish trait data (pipefishTraits.csv), output from docs/dataCleaning.Rmd and used in the downstream population analysis documented in <a href="https://github.com/spflanagan/ecology-duvauchelle/blob/main/docs/populationAnalyses.Rmd">docs/populationAnalyses.Rmd</a>.

opencc-by-4.0Sep 2023View details →

Genetic diversity, population structure, and linkage disequilibrium among tropical quality protein maize (QPM) lines assessed with high-density SNP markers

The study of genetic diversity (GD), population structure, and linkage disequilibrium (LD) provides a better understanding of the genetic relationships between individuals in a population which can be utilized in crop research and improvement. Genotyping-by-sequencing (GBS) was used to detect and genotype single nucleotide polymorphisms (SNPs) in a collection of 74 quality protein maize (QPM) lines and further to characterize their genetic diversity, population structure, and linkage disequilibrium. A total of 235,214 high-quality SNPs were used for different genetic analyses except for structure analysis where 11,950 SNPs were used. Analysis of molecular variance (AMOVA) based on these SNPs revealed high genetic heterozygosity among the five populations with 1% of the total genetic variation present among the subpopulations and 99% of the variation among individuals within the populations.  Population structure analysis using Bayesian-based clustering revealed that the 74 lines could be clustered into four groups. However, neighbor-joining trees indicate the lines are grouped into three major clusters.  Further analysis using principal component analyses (PCA) clustered the genotypes into five groups which are concordant with the groups based on pedigree information. Higher genetic diversity was detected in population 1 with a GD value of 0.484 and the lowest in population 5 (0.396) and overall, with a mean of 0.434. The LD pattern in the quality protein maize was investigated and we observed a relatively rapid LD decay of 3.53kb and 10.66kb at r2 =0.2 and r2= 0.1, respectively. Our findings provide important information for future Linkage mapping studies, genome-wide association analyses, and marker-assisted selective breeding of maize as well as genomic prediction-based selection in tropical germplasm.

opencc-zeroJan 2020View details →

Geographical gradients of genetic diversity and differentiation among the southernmost marginal populations of Abies sachalinensis revealed by EST-SSR polymorphism

Research Highlights: We detected the longitudinal gradients of genetic diversity parameters, such as the number of alleles, effective number of alleles, heterozygosity, and inbreeding coefficient, and found that these might be attributable to climatic conditions, such as temperature and snow depth. Background and Objectives: Genetic diversity among local populations of a plant species at its distributional margin has long been of interest in ecological genetics. Populations at the distribution center grow well in favorable conditions, but those at the range margins are exposed to unfavorable environments, and the environmental conditions at establishment sites might reflect the genetic diversity of local populations. This is known as the central-marginal hypothesis in which marginal populations show lower genetic variation and higher differentiation than do central populations. In addition, genetic variation in a local population is influenced by phylogenetic constraints and the population history of selection under environmental constraints. In this study, we investigated this hypothesis in relation to Abies sachalinensis, a major conifer species in Hokkaido. Materials and methods: A total of 1,189 trees from 25 natural populations were analyzed using 19 EST-SSR loci. Results: The eastern populations; namely, those in the species distribution center, showed greater genetic diversity than did the western peripheral populations. Another important finding is that the southwestern marginal populations were highly differentiated from the other populations. Conclusions: These differences might be due to genetic drift in the small and isolated populations at the range margin. Therefore, our results indicated that the central-marginal hypothesis held true for the southernmost A. sachalinensis populations in Hokkaido.

opencc-by-4.0Apr 2019View details →

Fig. 3 in Genetic diversity and population structure of Brycon nattereri (Characiformes: Bryconidae): a Neotropical fish under threat of extinction

Fig. 3. Haplotype network based on partial sequencing of the D-loop region (mtDNA) of 92 individuals of Brycon nattereri from the Laranjinha River. Circle sizes are pro- portional to haplotype frequency.

dryad40/100

Data from: Drift happens: molecular genetic diversity and differentiation among populations of jewelweed (Impatiens capensis Meerb.) reflect fragmentation of floodplain forests

Landscape features often shape patterns of gene flow and genetic differentiation in plant species. Populations that are small and isolated enough also become subject to genetic drift. We examined patterns of gene flow and differentiation among 12 floodplain populations of the selfing annual jewelweed (Impatiens capensis Meerb.) nested within four river systems and two major watersheds in Wisconsin, USA. Floodplain forests and marshes provide a model system for assessing the effects of habitat fragmentation within agricultural/urban landscapes and for testing whether rivers act to genetically connect dispersed populations. We generated a panel of 12,856 single nucleotide polymorphisms and assessed genetic diversity, differentiation, gene flow, and drift. Clustering methods revealed strong population genetic structure with limited admixture and highly differentiated populations (mean multilocus FST = 0.32, FST' = 0.33). No signals of isolation by geographic distance or environment emerged, but alleles may flow along rivers given that genetic differentiation increased with river distance. Differentiation also increased in populations with fewer private alleles (R2 = 0.51) and higher local inbreeding (R2 = 0.22). Populations varied greatly in levels of local inbreeding (FIS = 0.2 to 0.9) and FIS declined in smaller, more isolated populations. These results suggest that genetic drift dominates other forces in structuring these Impatiens populations. In rapidly changing environments, species must migrate or genetically adapt. Habitat fragmentation limits both processes, potentially compromising the ability of species to persist in fragmented landscapes.

opencc-zeroDec 2018View details →

FIGURE 30 in Morphological and molecular data reveal the cryptic diversity among populations of Aegla paulensis (Decapoda, Anomura, Aeglidae), with descriptions of four new species and comments on dispersal routes and conservation status

FIGURE 30. Bayesian tree (TPM 2 uf + G) for Aegla species based on partial fragment of 16 S. Node numbers represent posterior probabilities (values <50 % are not shown), and divergence time in millions of years (my); * indicates the calibration points to molecular clock. The clade C proposed by Pérez-Losada et al. (2004) is highlighted in grey. The basin and sub-basin origin of the discussed species in this study are shown after the specific names.

FIGURE 24. A – L in Morphological and molecular data reveal the cryptic diversity among populations of Aegla paulensis (Decapoda, Anomura, Aeglidae), with descriptions of four new species and comments on dispersal routes and conservation status

FIGURE 24. A – L, proximal portion of fifth pereiopod showing coxa and sexual tube of long and narrow type. A – B, Aegla paulensis Schmitt, 1942 s. str., male topotype (MZUSP 34368). C – D, Aegla rosanae Campos Jr., 1998, male topotype (MZUSP 34369). E – F, Aegla vanini n. sp., male paratype (MZUSP 34372). G – H, Aegla japi n. sp., male paratype (MZUSP 34375). I – J, Aegla jaragua n. sp. male paratype (MZUSP 34378). K-L, Aegla jundiai n. sp., male paratype (MZUSP 13490). Bars: A – D, F – H, J = 200 µm; K, L = 100 µm; E, I = 500 µm.

FIGURE 8 in Morphological and molecular data reveal the cryptic diversity among populations of Aegla paulensis (Decapoda, Anomura, Aeglidae), with descriptions of four new species and comments on dispersal routes and conservation status

FIGURE 8. Types of Aegla Leach, 1820 male sexual tubes. A, long and narrow (A. lancinhas Bond-Buckup & Buckup in Santos et al., 2015, MZUSP 34403). B, short and wide (A. leptochela Bond-Buckup & Buckup, 1994, MZUSP 34491).

FIGURE 1 in Morphological and molecular data reveal the cryptic diversity among populations of Aegla paulensis (Decapoda, Anomura, Aeglidae), with descriptions of four new species and comments on dispersal routes and conservation status

FIGURE 1. Distribution of the species of Aegla in four main hydrographic basins of southern Brazil: Rio Grande, Rio Tietê (Upper Paraná system), Rio Paraíba do Sul and Ribeira de Iguape. Indications L 1 through L 7 refer to the locations mentioned under “ sampling area ” in the Material & Methods section.