Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
13,064
datasets available to search
ShareScore release 0.9.0
Dataset results
13,064 results for “Prediction”
Social networks predict the life and death of honey bees - Data
<p><strong>Interaction matrices and metadata used in "Social networks predict the life and death of honey bees"</strong></p> <p><a href="https://www.biorxiv.org/content/10.1101/2020.05.06.076943v2">Preprint: Social networks predict the life and death of honey bees</a></p> <p>See the README file in <a href="https://doi.org/10.5281/zenodo.4435058">bb_network_decomposition</a> for example code.</p> <p><strong>The following files are included:</strong></p> <p><strong>interaction_networks_20160729to20160827.h5</strong></p> <p>The social interaction networks as a dense tensor and metadata.</p> <p>Keys:</p> <ul> <li>interactions: Tensor of shape (29, 2010, 2010, 9) (days x individuals x individuals x interaction_types). I_{d,i,j,t} = log(1 + x), where x is the number of interactions of type t between individuals i and j at recording day d. See the methods section of paper of the interaction types.</li> <li>labels: Names of the 9 interaction types in the order they are stored in the interactions tensor.</li> <li>bee_ids: List of length 2010, mapping from sequential index used in the interaction tensor to the original BeesBook tag ID of the individual</li> </ul> <p><strong>alive_bees_bayesian.csv </strong></p> <p>This file contains the results of the bayesian lifetime model with one row for each bee.</p> <p>Columns:</p> <ul> <li>bee_id: Numerical unique identifier for each individual.</li> <li>days_alive: Number of bees the bees was determined to be alive. If the individual was still alive at the end of the recording, the number of days from the day she hatched until the end of the recording.</li> <li>death_observed: Boolean indicator whether the death occurred during the recording period.</li> <li>annotated_tagged_date: Hatch date of the individual, i.e. the date she was tagged.</li> <li>inferred_death_date: The death date as determined by the model.</li> </ul> <p><strong>bee_daily_data.csv</strong></p> <p>This file contains one row per bee per day that she was alive for the focal period.</p> <p>Columns:</p> <ul> <li>bee_id: Numerical unique identifier for each individual.</li> <li>date: Date in year-month-day format.</li> <li>age: Age in days. Can be NaN if the bee has no associated death_date.</li> <li>network_age, network_age_1, network_age_2: The first three dimensions of network age.</li> <li>dance_floor, honey_storage, near_exit, brood_area_total: Normalized (sum to 1). Can be NaN if a bee had no high confidence detections (>0.9) for a given day. Can be 0 if a bee was only seen outside of the annotated areas.</li> <li>location_descriptor_count: The number of minutes the bee was seen in one of the location labels during that day. I.e., dance_floor * location_descriptor_count calculates the number of minutes, the bee was seen on the dance floor on the given day.</li> <li>death_date: Date the bee was last seen in the colony in year-month-day format. Can be NaN for individuals that did not die until the end of the recording period.</li> <li>circadian_rhythm: R² value of a sine with a period of one day fitted to the velocity data of the individual over three days. Can be NaN if the fit did not converge due to a lack of data points.</li> <li>velocity_peak_time: Phase of the circadian sine fit in hours as an offset to 12:00 UTC. Can be NaN if circadian_rhythm is NaN.</li> <li>velocity_day, velocity_night: Mean velocity of the individual between 09:00-18:00 UTC and 21:00-06:00 UTC, respectively. Can be NaN if no velocity data was available for that interval.</li> <li>days_left: Difference in days between date and death_date. Can be NaN if death_date is NaN.</li> </ul> <p><strong>location_data.csv</strong></p> <p>This file contains subsampled position information for all bees during the focal period. The data contains one row for every individual for every minute of the recording if that individual was seen at least once during that minute with a tag confidence of at least 0.9. The first matching detection for each individual is used.</p> <p>Columns:</p> <p>In addition to the bee_id and date columns as in the bee_daily_data.csv, the file contains these additional columns:</p> <ul> <li>cam_id, cams: The cam_id is a numerical identifier from {0, 1, 2, 3}. Each side of the hive is filmed by two cameras where {0, 1} and {2, 3} record the same side respectively. The cams column contains values either “(0, 1)” or “(2, 3)” and indicates to which sides of the hive this detection belongs.</li> <li>x_pos_hive, y_pos_hive: The spatial positions in millimeters on the hive. The two cameras from one side share a common coordinate system.</li> <li>location: The label that was assigned to the comb at (x_pos_hive, y_pos_hive) on the given date. The label “other” indicates detections that were outside of any annotated region. The label “not_comb” indicates the wooden frame or empty space around the comb.</li> <li>timestamp, date: The timestamp indicates the beginning of each one-minute sampling interval and is given in UTC, as indicated (example: “2016-08-13 00:00:00+00:00”). The date part of the timestamp is repeated in the “date” column. Both are given in year-month-day format.</li> </ul> <p><strong>Software used to acquire and analyze the data:</strong></p> <ul> <li><a href="https://doi.org/10.5281/zenodo.4435058">bb_network_decomposition: Network age calculation and regression analyses</a></li> <li><a href="https://github.com/BioroboticsLab/bb_pipeline/releases/tag/2016">bb_pipeline: Tag localization and decoding pipeline</a></li> <li><a href="https://github.com/BioroboticsLab/bb_pipeline_models/releases/tag/2016">bb_pipeline_models: Pretrained localizer and decoder models for bb_pipeline</a></li> <li><a href="https://github.com/BioroboticsLab/bb_binary/releases/tag/2016">bb_binary: Raw detection data storage format</a></li> <li><a href="https://doi.org/10.5281/zenodo.4436419">bb_irflash: IR flash system schematics and arduino code</a></li> <li><a href="https://github.com/BioroboticsLab/bb_imgacquisition/releases/tag/2016">bb_imgacquisition: Recording and network storage </a></li> <li><a href="https://github.com/BioroboticsLab/bb_behavior/releases/tag/2016">bb_behavior: Database interaction and data (pre)processing, velocity calculation</a></li> <li><a href="https://github.com/BioroboticsLab/bb_circadian/releases/tag/2016">bb_circadian: Circadian rhythm calculations</a></li> <li><a href="https://github.com/BioroboticsLab/bb_tracking_2016/releases/tag/2016">bb_tracking: Tracking of bee detections over time</a></li> <li><a href="https://github.com/BioroboticsLab/bb_wdd/releases/tag/2016">bb_wdd: Automatic detection and decoding of honey bee waggle dances</a></li> <li><a href="https://github.com/BioroboticsLab/bb_interval_determination/releases/tag/2016">bb_interval_determination: Homography calculation</a></li> <li><a href="https://github.com/BioroboticsLab/bb_stitcher/releases/tag/2016">bb_stitcher: Image stitching</a></li> </ul> <p> </p>
Dipole localisation predictions data set
<p>This data set contains prediction of ten dipole localisation algorithms computed using a simulated artificial lateral line (potential flow).</p>
Dataset for: Predicting stable lithium iron oxysulphides for battery cathodes
<p>Cathode materials that have high specific energies and low manufacturing costs are vital for the scaling up of lithium-ion batteries (LIBs) as energy storage solutions. Fe-based intercalation cathodes are highly attractive because of the low-cost and the abundance of the raw materials. However, existing Fe-based materials, such as LiFePO<sub>4</sub> suffer from low capacity due to the large size of the polyanions. Turning to mixed anion systems can be a promising strategy to achieve higher specific capacity. Recently, anti-perovskite structured oxysulphide Li<sub>2</sub>FeSO has been synthesised and reported to be electrochemically active.<br> In this work, we perform an extensive computational search for iron-based oxysulphides using <em>ab initio</em> random structure searching (AIRSS). By performing an unbiased sampling of the Li-Fe-S-O chemical space, several new oxysulphide phases have been discovered which are predicted to be less than 50 meV/atom from the convex hull and potentially accessible for synthesis.<br> Among the predicted phases, two anti-Ruddlesden-Popper structured materials Li<sub>2</sub>Fe<sub>2</sub>S<sub>2</sub>O and Li<sub>4</sub>Fe<sub>3</sub>S<sub>3</sub>O<sub>2</sub><br> have been found to be attractive as they have high theoretical capacities with calculated average voltages 2.9 V and 2.5 V respectively. With band gaps as low as about 2.0 eV, they are expected to exhibit good electronic conductivities.<br> By performing nudged-elastic band calculations, we show that the Li-ion transport in these materials takes place by hopping between the nearest neighbouring sites with low activation barriers between 0.3 eV and 0.5 eV.<br> The richness of new materials yet to be synthesised in the Li-Fe-S-O phase field illustrate the great opportunity in these mixed anion systems for energy storage applications and beyond.</p> <p> </p> <p>The dataset includes the structure searching results and outputs of further property calculations. The analysis codes are also included as Jupyter Notebooks.</p> <p> </p> <p>Also hosted on <a href="https://github.com/SMTG-UCL/Li-Fe-S-O-oxysulphides">GitHub</a>.</p> <p>Preprint hosted on <a href="https://doi.org/10.33774/chemrxiv-2021-fbffd-v2">ChemRxiv</a>.</p>
DWCox: A Density-Weighted Cox Model for Outlier-Robust Prediction of Prostate Cancer Survival
<p>This package, <strong>DWCox</strong>, implements a <strong>d</strong>ensity-<strong>w</strong>eighted <strong>Cox</strong> regression model that is more robust against outliers in the training data. DWCox gives more accurate predictions than the standard Cox regression on prostate cancer survival, especially in cases where the training data are expected to contain a lot of outliers. More details can be found in our paper (coming soon) and the README file inside this package.</p>
Dataset: Neural correlates of error prediction in a complex motor task
<p>There are two files for each subject:</p> <p>1. errorsegments_sub##.mat -> Contains EEG Segments, that were recorded while the subject executed a clear target miss (minimal distance between the center of the ball and target > 12 cm) in the task (segment and electrode information can be found below).</p> <p>2. hitsegments_sub##.mat -> Contains EEG Segments, that were recorded while the subject executed a clear target hit (minimal distance between the center of the ball and the target < 5 cm) in the task (segment and electrode information can be found below).</p> <p>The data in the *.mat-files are stored in a three dimensional matrix: 1300 datapoints x n segments x 14 electrodes</p> <p>datapoints: The first dimension contains 1300 data points for each segment which translates to 2600 ms (500 Hz sampling frequency). The time of the ball´s release set at the 301st data point in each segment.</p> <p>segments: The second dimension stands for the number of segments. Since the number of trials which satisfy the above described distance criterion for hit and error trials differ for participants size, n is variable. </p> <p>electrodes: The third dimension consists of the 14 different electrodes that were used during data recording in this exact order: [F3 Fz F4 C4 Cz C3 P3 Pz P4 VEOGu VEOGo HEOGre HEOGli FCz]</p> <p> </p> <p> </p>
Dataset: Brain negativity as an indicator of predictive error processing: The contribution of visual action effect monitoring
<p>There are two files for each subject:</p> <p>1. sub##_error.dat -> Contains EEG Segments, that were recorded while the subject executed a clear target miss (minimal distance between the center of the ball and target > 12 cm) in the task (segment and electrode information can be found below).</p> <p>2. sub##_hit.dat -> Contains EEG Segments, that were recorded while the subject executed a clear target hit (minimal distance between the center of the ball and the target < 7 cm) in the task (segment and electrode information can be found below).</p> <p><br> The data in the *.dat-files are stored in a two dimensional matrix: n*1400 datapoints x 15 electrodes</p> <p>n represents the number of segments. 1400 datapoints per segment translate to a segment length of 2800 ms (from 600 ms before to 2200 ms after ball release). The ball´s release is located at the 301st datapoint and the feedback was presented at datapoint 726 (850 ms after ball release) in every segment.</p> <p>datapoints: The first dimension (rows) includes the measured neural activations in microvolts. The data is stored vectorized,<br> i.e. hit/error #1 -> row 1 to 1400, hit/error #2 -> row 1401 to 2800, ..., hit/error #n -> (n-1) * 1400 + 1 to n * 1400</p> <p>electrodes: The second dimension (columns) consists of the 15 different electrodes that were used during data recording in this exact order: [F3 Fz F4 C4 Cz C3 P3 Pz P4 VEOGu VEOGo HEOGre HEOGli FCz Mastre]</p>
Vertical Profiles of Convection-Permitting Simulations for Predicting Thunderstorm Occurrence
<p>This repository contains datasets for training and evaluation of the machine learning (ML) models in K. Vahid Yousefnia et al., <em>Inferring Thunderstorm Occurrence from Vertical Profiles of Convection-Permitting Simulations: Physical Insights from a Physical Deep Learning Model</em>, 2024 (submitted to <em>Artificial Intelligence for the Earth Systems,</em> preprint available at https://arxiv.org/abs/2409.20087).</p>
Fully differentiable, fully distributed River Discharge Prediction: data sets
<p>This repository contains the data sets used in: Scholz et al. (2025). Fully differentiable, fully distributed River Discharge Prediction.</p> <ul> <li><code>dem_1000.h5</code> based on EU-DEM v1.1, reprojected to RADOLAN grid: <a href="https://sdi.eea.europa.eu/catalogue/srv/api/records/3473589f-0854-4601-919e-2e7dd172ff50">https://sdi.eea.europa.eu/catalogue/srv/api/records/3473589f-0854-4601-919e-2e7dd172ff50</a></li> <li><code>efas.h5</code> based on EFAS historical: <a href="https://ewds.climate.copernicus.eu/datasets/efas-historical?tab=overview">https://ewds.climate.copernicus.eu/datasets/efas-historical?tab=overview</a></li> <li><code>era5_ssrd_neckar*.nc</code> based on ERA5 provided by ECMWF, reprojected to RADOLAN grid: <a href="https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5">https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5</a></li> <li><code>radolan_neckar_*.h5</code> based on RADOLAN rw product provided by the Deutsche Wetterdienst: <a href="https://opendata.dwd.de/climate_environment/CDC/grids_germany/hourly/radolan/">https://opendata.dwd.de/climate_environment/CDC/grids_germany/hourly/radolan/</a></li> </ul> <p>Due to copyright, the discharge data has to be downloaded manually from the Global Runoff Data Centre (<a href="https://grdc.bafg.de/">https://grdc.bafg.de/</a>), and then preprocessed with the provided <code>bafg_parser.py</code> python script. We use the following stations in our work:</p> <ul> <li>6335290: STEIN</li> <li>6335291: GAILDORF</li> <li>6335565: BAD IMNAU</li> <li>6335600: ROCKENAU SKA</li> <li>6335601: LAUFFEN</li> <li>6335602: PLOCHINGEN</li> <li>6335603: ROTTWEIL</li> <li>6335604: KIRCHENTELLINSFURT</li> <li>6335620: MOSBACH</li> <li>6335660: PFORZHEIM</li> <li>6335665: DENKENDORF</li> <li>6335671: ALTENSTEIG</li> <li>6335675: MURR</li> <li>6335676: OPPENWEILER</li> <li>6335680: SCHWABSBERG</li> <li>6335681: UNTERGRIESHEIM</li> <li>6335690: NEUSTADT</li> </ul> <p>To preprocess the discharge data, additionally the river network data "Fließgewässer (AWGN)" provided by the Landesanstalt für Umwelt Baden-Württemberg (LUBW) is required: <a href="https://rips-metadaten.lubw.de/trefferanzeige?docuuid=7251515f-6aed-4555-8319-ab6314155ab1">https://rips-metadaten.lubw.de/trefferanzeige?docuuid=7251515f-6aed-4555-8319-ab6314155ab1</a></p> <p> </p>
Link-prediction on Biomedical Knowledge Graphs
<p>Release of code and experimental data from the paper <em>Towards Linking Graph Topology to Model Performance for Biomedical Knowledge Graph Completion </em>(<em>Machine Learning for Life and Material Sciences</em> workshop @ ICML2024) and <a href="https://arxiv.org/abs/2409.04103" rel="nofollow">The Role of Graph Topology in the Performance of Biomedical Knowledge Graph Completion Models</a>.</p> <div> <div>Knowledge Graph Completion has been increasingly adopted as a useful method for several tasks in biomedical research, like drug repurposing or drug-target identification. To that end, a variety of datasets and Knowledge Graph Embedding models has been proposed over the years. However, little is known about the properties that render a dataset useful for a given task and, even though theoretical properties of Knowledge Graph Embedding models are well understood, their practical utility in this field remains controversial. We conduct a comprehensive investigation into the topological properties of publicly available biomedical Knowledge Graphs and establish links to the accuracy observed in real-world applications. By releasing all model predictions we invite the community to build upon our work and continue improving the understanding of these crucial applications.</div> <div> </div> <div>Experiments were conducted on six datasets: five from the biomedical domain (<a href="../records/268568">Hetionet</a>, <a href="https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/IXA7BM">PrimeKG</a>, <a href="../records/4077338">PharmKG</a>, <a href="../records/5361324">OpenBioLink2020 HQ</a>, <a href="../records/7011027">PharMeBINet</a>) and one trivia KG (<a href="https://aclanthology.org/W15-4007.pdf">FB15k-237</a>). All datasets were randomly split into training, validation and test set (80% / 10% / 10%; in the case of PharMeBINet, 99.3% / 0.35% / 0.35% to mitigate the increased inference cost on the larger dataset).</div> <div>On each dataset, five different KGE models were compared: <a href="https://dl.acm.org/doi/10.5555/2999792.2999923">TransE</a>, <a href="https://arxiv.org/abs/1412.6575">DistMult</a>, <a href="https://arxiv.org/abs/1902.10197">RotatE</a>, <a href="https://arxiv.org/abs/2209.08271">TripleRE</a>, <a href="https://dl.acm.org/doi/10.5555/3504035.3504256">ConvE</a>. Hyperparameters were tuned on the validation split (see final train configurations in <code>train/scripts</code>). We release results for tail predictions on the test split. In particular, each test query <code>(h,r,?)</code> is scored against all entities in the KG and we compute the rank of the score of the correct completion <code>(h,r,t)</code> , after masking out scores of other <code>(h,r,t')</code> triples contained in the graph.</div> <div>Note: the ranks provided are computed as the average between the optimistic and pessimistic ranks of triple scores.</div> <div> </div> <div>Inside <code>experimental_data.zip</code>, the following files are provided.</div> <div> <ul> <li><code>datasets/{dataset}</code>: a folder for each dataset, containing <ul> <li><code>{dataset}_preprocessing.ipynb</code>: a Jupyter notebook for downloading and preprocessing the datasets. In particular, this generates the custom label->ID mapping for entities and relations, and the numerical tensor of <code>(h_ID,r_ID,t_ID)</code> triples for all edges in the graph, which can be used to compute graph topological metrics (e.g., using <a href="https://github.com/graphcore-research/kg-topology-toolbox">kg-topology-toolbox</a>) and compare them with the edge prediction accuracy.</li> <li><code>test_ranks.csv</code>: csv table with columns <code>["h", "r", "t"]</code> specifying the head, relation, tail IDs of the test triples, and columns <code>["DistMult", "TransE", "RotatE", "TripleRE", "ConvE"]</code> with the rank of the ground-truth tail in the ordered list of predictions made by the five KGE models;</li> <li><code>entity_dict.csv</code>: list of entity labels, ordered by entity ID (as generated in the preprocessing notebook);</li> <li><code>relation_dict.csv</code>: list of relation labels, ordered by relation ID (as generated in the preprocessing notebook).</li> </ul> </li> <li><code>train</code>: code to reproduce training (and validation) of the five KGE models, using the <a href="https://github.com/graphcore-research/bess-kge">BESS-KGE</a> distribution framework. <ul> <li><code>train/scripts</code>: executable scripts, with specifications of the final hyperparameters for all models and datasets.</li> </ul> </li> <li><code>notebooks</code>: Jupyter notebooks for data analysis and generation of all the figures in the paper.</li> </ul> <p>The separate <code>top_100_tail_predictions.zip</code> archive contains, for each of the test queries in the corresponding <code>test_ranks.csv</code> table, the IDs of the top-100 tail predictions made by each of the five KGE models, ordered by decreasing likelihood. The predictions are released in a <code>.npz</code> archive of numpy arrays (one array of shape <code>(n_test_triples, 100)</code> for each of the KGE models). </p> </div> </div>
PubChemLite for Exposomics + predicted CCS from CCSbase - 5 Sept. 2025
<p>PubChemLite is a subset of PubChem (<a href="https://pubchem.ncbi.nlm.nih.gov/">https://pubchem.ncbi.nlm.nih.gov/</a>) selected from major categories of the Table of Contents page at the PubChem Classification Browser (<a href="https://pubchem.ncbi.nlm.nih.gov/classification/#hid=72">https://pubchem.ncbi.nlm.nih.gov/classification/#hid=72</a>). This version of PubChemLite for Exposomics has predicted collision cross section (CCS) values for 11 adducts provided by Libin Xu and team at CCSbase (<a href="https://ccsbase.net/">https://ccsbase.net/</a>) calculated from the latest <a href="https://github.com/dylanhross/c3sdb/">c3sdb</a> code.</p> <p>PubChemLite <em>exposomics</em> is compiled from 11 categories: AgroChemInfo, BioPathway, DrugMedicInfo, FoodRelated, PharmacoInfo, SafetyInfo, ToxicityInfo, KnownUse, DisorderDisease, Identification, ChemClass.</p> <p>CCS adducts provided are: [M+H]+, [M-H]-, [M+Na]+, [M+K]+, [M+NH4]+, [M+H-H2O]+, [M+HCOO]-, [M+CH3COO]-, [M+Na-2H]-, [M]+, [M]-</p> <p>Details on the CCS prediction are given here: Ross <em>et al</em>. (2020) Analytical Chemistry, DOI: <a href="https://pubs.acs.org/doi/10.1021/acs.analchem.9b05772">10.1021/acs.analchem.9b05772</a></p> <p>PubChemLite is described in Schymanski <em>et al. </em>(2021) J. Cheminformatics, DOI: <a href="https://doi.org/10.1186/s13321-021-00489-0" target="_blank" rel="noopener">10.1186/s13321-021-00489-0</a></p> <p>An article describing these joint efforts is available: Elapavalore <em>et al</em>. (2025) ES&T Letters, DOI: <a href="https://doi.org/10.1021/acs.estlett.4c01003">10.1021/acs.estlett.4c01003</a></p> <p>PubChemCIDs have been collapsed by InChIKey first block, reporting the structure from the most annotated CID, plus related CIDs. Entries that will be ignored by MetFrag (salts, disconnected substances) or cause errors (e.g. transition metals) have been removed. The Patent and PubMed ID counts are extracted from files on the PubChem FTP site. The "AnnoTypeCount" term counts how many of the categories are represented, the subsequent column (named per category) counts the number of annotation categories available in the next sub-category of the TOC entry.</p> <p>These files can be used "as is" as localCSV for MetFrag Command Line (<a href="https://ipb-halle.github.io/MetFrag/">https://ipb-halle.github.io/MetFrag/</a>) - please do NOT upload these files directly to the web interface, they are too large and will be available in a drop-down menu.</p> <p>Further details are described in Schymanski <em>et al.</em> (2021) DOI:<a href="https://doi.org/10.1186/s13321-021-00489-0">10.1186/s13321-021-00489-0</a> and Elapavalore <em>et al</em>. (2025) DOI: <a href="https://doi.org/10.1021/acs.estlett.4c01003">10.1021/acs.estlett.4c01003</a></p> <p><strong><em>NOTE: The latest PubChemLite for Exposomics version can be downloaded at DOI:</em></strong><em><strong><a href="https://doi.org/10.5281/zenodo.5995885">10.5281/zenodo.5995885</a> (currently updating monthly). This file will be updated shortly after. <br></strong></em></p> <p>Please cite this data source and Elapavalore <em>et al</em>. (2025) DOI: <a href="https://doi.org/10.1021/acs.estlett.4c01003">10.1021/acs.estlett.4c01003</a> when using this dataset.</p>
Numerically predicted permeability of over 6500 artificially generated fibrous microstructures
<p>This data set was generated in the project "ML4ProcessSimulation - Machine Learning for Simulation Intelligence in Composite Process Design" (Leibniz Collaborative Excellence funding program: K377/2021), at Leibniz-Institut für Verbundwerkstoffe GmbH. The goals were to create a comprehensive data set for training different neural networks and to gain insight into the influence of fiber structure on permeability. The models represent the fiber structure within fiber bundles in fiber-reinforced plastic composites (FRPC). Over 6500 structure models were generated in the software GeoDict® [1] and the permeability tensor of these models was numerically calculated in the GeoDict® module FlowDict [2]. The zip files contain the structure file (<i>gdt</i>), the model generation result file (<i>FiberGeo_[...].gdr</i>) and the flow simulation result file (<i>LIRStokesResult_[...].gdr</i>). For each zip file is a JSON meta data file available and in addition the gdr files contain all input and output data of the model generation and the flow simulation. The file <i>Table_of_Parameter_studies_and_model_pictures.jpg</i> gives an overview of the parameter studies and exemplarily shows two models each.The data set is divided into three parameter studies: </p><ul><li>1_Parameter_study_round_fibers with round fibers by varying the fiber volume content (fvc), fiber diameter (fdia) and fiber orientation (fdir). For each modeling parameter, 5 - 100 models (random seed or RS) were generated, all differing due to the randomized fiber positioning during model generation.</li><li>2_Parameter_study_elliptical_fibers with elliptical fibers that was varied based on different aspect ratios (asp1, asp2, asp3). In addition, fdia and fvc were varied and 5 models (RS) were calculated. </li><li>3_Parameter_study_undulation with elliptical fibers, whose undulation was varied. In addition, fdia and fvc were varied and 12 models (RS) were calculated.</li></ul><p><i>[1] J. Hilden, S. Rief, and B. Planas, GeoDict 2023 User Guide. FiberGeo handbook. DE: Math2Market GmbH, 2023. Accessed: Oct. 26, 2023. [Online]. Available: https://doi.org/10.30423/userguide.geodict</i></p><p><i>[2] J. Hilden, S. Linden, and B. Planas, "GeoDict 2023 User Guide. FlowDict handbook." Math2Market GmbH, 2023. Accessed: Jul. 31, 2023. [Online]. Available: https://doi.org/10.30423/userguide.geodict</i></p>
Datasets of sequences, alignments and structural models generated for the structural prediction of complexes mediated by intrinsically disordered regions.
<p>This repository contains input and ouput files used and generated for the scanning of intrinsically disordered region and the prediction of their binding sites to receptor proteins using the <a href="https://github.com/i2bc/SCAN_IDR">SCAN_IDR</a> pipeline with AlphaFold2-Multimer.</p><p>It contains two archives: </p><ol><li><a href="https://zenodo.org/api/records/10068949/draft/files/scanidr_data_repository_corr6J08.tar/content"><i><strong>scanidr_data_repository_corr6J08.tar</strong></i></a> dedicated to the analysis of a dataset of 42 protein complexes non redundant with the dataset used for AlphaFold2 training,</li><li><a href="https://zenodo.org/api/records/10068949/draft/files/923_elm_cases_repository.tar.gz/content"><i><strong>923_elm_cases_repository.tar.gz</strong></i></a> dedicated to the analysis of 923 complexes from the ELM database.</li></ol><p>These data can be used to rerun specific sections of the pipeline and scripts provided in: <a href="https://github.com/i2bc/SCAN_IDR">https://github.com/i2bc/SCAN_IDR</a></p><h4><strong>Dataset of 42 non redundant complexes</strong></h4><p>The first archive <a href="https://zenodo.org/api/records/10068949/draft/files/scanidr_data_repository_corr6J08.tar/content"><i><strong>scanidr_data_repository_corr6J08.tar</strong></i></a> contains 3 compressed directories and a README file detailing their contents :</p><ul><li>the initial raw sequence and alignment data for every chain -> DIRECTORY <strong>fasta_msa/</strong></li><li>the input and output data of every Alphafold run for every complex -> DIRECTORY <strong>af2_runs/</strong></li><li>the native reference structures -> DIRECTORY <strong>ref_capri_curated/</strong></li></ul><p>The protein-peptide complex cases have been assigned a distinct index number, from 1 to 42, consistent across the several directories of the archive. Their corresponding directories are labelled as <i><index>_<pdbcode></i>.</p><p><i>The models in this archive were generated using AlphaFold2-Multimer v2.2</i></p><h4><strong>Dataset of 923 complexes selected from the ELM database</strong></h4><p>The second archive <a href="https://zenodo.org/api/records/10068949/draft/files/923_elm_cases_repository.tar.gz/content"><i><strong>923_elm_cases_repository.tar.gz</strong></i></a> contains input and ouput files used and generated for the analysis of 923 Eukaryotic Linear Motifs (ELM) database entries.</p><p>Each ELM entry is indexed with specific integer id and is composed of a receptor and a ligand protein. </p><p>The archive contains a Table associating ELM indexes with the ELM entry information, 5 directories and a README file detailing their contents:</p><ul><li>the table describing ELM entries -> FILE <strong>Table_923ELM_uid_delimitations_info_for_archive.txt</strong></li><li>the initial raw sequence and multiple sequence alignment (MSA) data for every chain -> DIRECTORY <strong>fasta_msa/</strong></li><li>the concatenated MSA model for every ELM complex and protocol used -> DIRECTORY <strong>af2_elm_coali_inputs/</strong></li><li>the best model of every AF2 protocol for every complex according to the AF2 -> DIRECTORY <strong>af2_elm_models/</strong></li><li>the best model cut in the ligand part to select only the ELM motifs as used for the evaluation of the models -> DIRECTORY <strong>elm_cut_models/</strong></li><li>the reference structures used for the evaluation of the models -> DIRECTORY <strong>ref_capri_curated/</strong></li></ul><p><i>The models in this archive were generated using AlphaFold2-Multimer v2.3</i></p>
Predicting Phenotypic Traits Using a Massive RNA-seq Dataset
<h2><strong>Abstract</strong></h2><p>The included datasets are a conglomerate of all available <i>Arabidopsis thaliana</i> RNA-seq data available from NCBI as of November 2022 processed to count data. In addition, the associated annotation files from NCBI BioProject database and processed versions of this data is included. Data has been processed according to the "Data Description Methods" in the manuscript titled "Predicting Phenotypic Traits Using a Massive RNA-seq Dataset" (in publication). The associated Methods can be found at this repository:<a href="https://gitlab.com/ficklinlab-public/modeling-with-transcriptomics"> https://gitlab.com/ficklinlab-public/modeling-with-transcriptomics</a>. These datasets can be used for exploring machine learning methods for predicting both continuous (Age) and categorical (Tissue) phenotypic traits using gene expression. Additionally, the gene expression data can be used on its own for the investigation of gene expression in <i>Arabidopsis thaliana.</i></p><h3><strong>Note to Researchers</strong></h3><p>This repository contains all of the datasets and information necessary to recreate the experiments in our paper. However, if may be that you are interested in our dataset for testing your own hypotheses/programs. If this is the case, we predict that you are looking for one or more of the the following 5 datasets<br> </p><h3><strong>Note on File Compression</strong></h3><p>All files in this repository are compressed using bzip2 to conserve space and allow for easier file transfer. The unzip command on linux systems is `bzip2 -d FILE_NAME`. For other computer systems (Windows and Apple) please consult your user manual.</p><h3><strong>Description All Datasets:</strong></h3><p><strong>Title:</strong> Gene Expression Count Data of all <i>Arabidopsis thaliana</i> data available from NCBI SRA as of November 2022<br><strong>Abstract: </strong>Gene Expression Count data was created using the workflow GEMmaker. The resulting Gene Expression Matrix (GEM) was then normalized and thresholded. The following 4 files are normalizations of the same data for Trimmed Mean of M values (TMM), Median Ratios Normalization (MRN), Transcripts Per kilobase Million (TPM), and No Normalization (NoNo) respectively. Additionally, Each file is included as a tsv and a python pickle. The tsv file is human readable, whereas the pickle file can be read into memory substantially faster. Format for tsv is each row represents a sample and each column represents a gene. <strong>NCBI_Nov2022_SRR_runinfo.csv</strong> is the starting file from NCBI which reports SRR information for each sample. <strong>Note 1 to Researchers: </strong>MRN normalization performed the best in our experiments and is likely what you want to use if you are doing additional expermentation with this dataset. Otherwise start with NoNo and perform your own normalizations. <strong>Note 2 to Researchers:</strong> the 54547 dataset will need to be thresholded prior to use. We include it in addition to the 32432 datasets in case you wish to try a different thresholding to the one outlined in our manuscript. <br><strong>Author:</strong> John Anthony Hadish<br><strong>Data Type: </strong>Gene Expression Count Data<br><strong>Organism:</strong> <i>Arabidopsis thaliana</i><br><strong>Files:</strong></p><p><strong>NCBI_Nov2022_SRR_runinfo.csv - </strong>Arabidopsis RNA-seq SRA RunInfo Retrieved from NCBI November 2022. This is the unprocessed data.<br><strong>Dataset_54547_NoFilter_raw.pkl </strong>- Raw File Before thresholding (".pkl" format). Same as NoNo normalization without thresholding.<br><strong>Dataset_54547_NoFilter_raw.tsv </strong>- Raw File Before thresholding (".tsv" format). Same as NoNo normalization without thresholding.<br><strong>Dataset_32432_MRN.pkl</strong> - MRN normalized (".pkl" format)<br><strong>Dataset_32432_MRN.tsv - </strong>MRN normalized (".tsv" format)<br><strong>Dataset_32432_NoNo.pkl - </strong>NoNo normalized (".pkl" format)<br><strong>Dataset_32432_NoNo.tsv - </strong>NoNo normalized (".tsv" format)<br><strong>Dataset_32432_TMM.pkl - </strong>TMM normalized (".pkl" format)<br><strong>Dataset_32432_TMM.tsv - </strong>TMM normalized (".tsv" format)<br><strong>Dataset_32432_TPM.pkl - </strong>TPM normalized (".pkl" format)<br><strong>Dataset_32432_TPM.tsv - </strong>TPM normalized (".tsv" format)<br><br><br><strong>Title: </strong>Meta Data Arabidopsis Age and Tissue<br><strong>Abstract: </strong>Meta Data for Age and Tissue after processing. In our experiment this was used as response variable to gene expression. Shared columns are "bio_sample", "bioproject_name", "experiment". In addition to these processed datasets, <strong>NCBI_Nov2022_BioSample_data.tsv </strong>is the unprocessed starting material for these two data frames.<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>Metadata on phenotypes. ".tsv" format <br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files: </strong><br><strong>NCBI_Nov2022_BioSample_data.tsv - </strong>Arabidopsis BioSample data retrieved from NCBI November 2022. This is the unprocessed data.<br><strong>df_metadata_tissue.tsv</strong> - Tissue Annotations for 24876 samples<br><strong>df_metadata_age.tsv</strong><i><strong> - </strong></i>Age Annotations for 16078 samples. In addition to shared columns includes<i> "</i>days<i>_</i>age"(how many days old the sample is converted to days) and "annotation_age" (how the annotation was reported for this sample in the raw data file-- i.e. "days", "weeks" etc.)<br><br><br><strong>Title: </strong>Machine Learning Dataset for <i>Arabidopsis thaliana</i> <strong>Age</strong><br><strong>Abstract: </strong>The dataset used for Machine learning on the phenotype Age that is a combination of the Gene Expression Matrix and the Annotation Matrix. Consists of a list of 4 for the train and test splits.<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>Gene Expression Matrix and Annotations Combined, split into train and test <br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files:</strong><br><strong>Dataset_Age_TrainTestSplits_mrn.pkl</strong> - MRN normalized<br><strong>Dataset_Age_TrainTestSplits_NoNo.pkl </strong>- NoNo normalized<br><strong>Dataset_Age_TrainTestSplits_tmm.pkl </strong>- TMM normalized<br><strong>Dataset_Age_TrainTestSplits_tpm.pkl - </strong>TPM normalized<br><br><br><strong>Title: </strong>Machine Learning Dataset for <i>Arabidopsis thaliana</i> <strong>Tissue</strong><br><strong>Abstract: </strong>The dataset used for Machine learning on the phenotype Tissue that is a combination of the Gene Expression Matrix and the Annotation Matrix. Consists of a list of 4 for the train and test splits. Saved as python ".pkl" files.<br><strong>Author:</strong> John Anthony Hadish<br><strong>Data Type: </strong>Gene Expression Matrix and Annotations Combined, split into train and test. Saved as python ".pkl" files.<br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files:</strong><br><strong>Dataset_Tissue_TrainTestSplits_mrn.pkl </strong>- MRN normalized<br><strong>Dataset_Tissue_TrainTestSplits_NoNo.pkl </strong>- NoNo normalized<br><strong>Dataset_Tissue_TrainTestSplits_tmm.pkl </strong>- TMM normalized<br><strong>Dataset_Tissue_TrainTestSplits_tpm.pkl </strong>- TPM normalized<br><strong>Dataset_Tissue_TrainTestSplits_mrn_4category.pkl </strong>- MRN for the tissue-4 dataset<br><br><br><strong>Title: </strong>BioProject Names<br><strong>Abstract:</strong> Three Column File With BioProject Name, BioSample Name, and Experiment Name<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>".tsv"<br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files:</strong><br><strong>BioProject_Names_All.tsv</strong><br><br><br><strong>Title:</strong> Manuscript Supplemental Material<br><strong>Abstract:</strong> Supplemental tables and figures described in the manuscript (included with manuscript and here for convenience). Please see manuscript for additional information.<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>".tsv", ".png".pdf"<br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files:</strong><br><strong>Supplemental_Figures.zip </strong>- Supplemental figures from the manuscript. Includes description of each figure.<br><strong>Supplemental_Tables.zip </strong>- Supplemental tables from the manuscript. Includes description of each table.<br><br> </p><p><strong>Title:</strong> Splits of data for 3 experiments<br><strong>Abstract:</strong> 2 column tsv files. The first column is the experiment (sample) name, and the second column is if it is included in the train or test data. <strong>Included here to make sure pkl files are reproducible in case the pkl package breaks in the future.</strong> Not used by scripts, included to prevent future potential loss of data.<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>".tsv"<br><strong>Organism: </strong><i>Arabidopsis thaliana</i><br><strong>Files:</strong><br><strong>Dataset_Tissue_TrainTestSplits_4category_namesOnly.tsv</strong><br><strong>Dataset_Tissue_TrainTestSplits_namesOnly.tsv</strong><br><strong>Dataset_Age_TrainTestSplits_namesOnly.tsv</strong></p><p> </p><p><strong>Title:</strong> Git Code Repository<br><strong>Abstract:</strong> A tar bz2 compression of the git repository containing all of the code created for this manuscript. The same code found in this file is also avalible on GitLab at the link: https://gitlab.com/ficklinlab-public/modeling-with-transcriptomics<br><strong>Author: </strong>John Anthony Hadish<br><strong>Data Type: </strong>Git Repository, python code<br><strong>Files:</strong><br><strong>modeling-with-transcriptomics-main.tar.bz2</strong> - Compressed Git repository of all code used in paper.</p>
Dataset and plot generation script for article "Probabilistic short-range forecasts of high precipitation events : optimal decision thresholds and predictability limits" by Francois Bouttier and Hugo Marchal, submitted in Dec 2023.
<p>Dataset and plot generation script for article "Probabilistic short-range forecasts of high precipitation events : optimal decision thresholds and predictability limits" by Francois Bouttier and Hugo Marchal, submitted in NHESS journal in Dec 2023.</p> <p>For further technical details read the file READMEdata in the zipfile. The script MAKEFIG remakes all the figures from the data.</p> <p>For scientific details read the associated article preprint on the NHESS egusphere website.</p>
Stiffness Moduli Modelling and Prediction in Four-Point Bending of Asphalt Mixtures: A Machine Learning-Based Framework within Weave-UNISONO 2021 project, NCN project No 2021/03/Y/ST8/00079, and GACR project GA22-04047K
<div><strong>Summary:</strong></div> <div>Two selected mixtures were thoroughly investigated in an experimental trial carried out by means of a four-point bending test (4PBT) apparatus. The mixtures were prepared using spilite aggregate, a conventional 50/70 penetration grade bitumen, and limestone filler. Their stiffness moduli (SM) were determined while samples were exposed to 11 loading frequencies (from 0.1 to 50 Hz) and 4 testing temperatures (from 0 to 30 °C). Observations were recorded and used to develop a machine learning (ML) model. The main scope was the prediction of the stiffness moduli based on the volumetric properties and testing conditions of the corresponding mixtures, which would provide the advantage of reducing the laboratory efforts required to determine them.</div> <div> </div> <div><strong>The dataset includes:</strong></div> <div>Characteristics of bituminous binder, CSV raw data</div> <div> <ul> <li>bituminous binder.csv</li> </ul> </div> <div>Grading curves of tested asphalt mixtures</div> <ul> <li>AML16 Grading curves.csv</li> <li>AMP22 Grading curves.csv</li> </ul> <div>Volumetric characterizations of AML16 and AMP22 mixtures</div> <ul> <li>AML16 Volumetric characterizations.csv</li> <li>AMP22 Volumetric characterizations.csv</li> </ul> <div>Outcomes of the 4PBT experimental trial carried out on AML16 and AMP22 mixtures</div> <ul> <li>AML16 Stiffness Modulus 4PB.csv</li> <li>AMP22 Stiffness Modulus 4PB.csv</li> </ul>
NPClassifier predictions of COCONUT compounds
<p>Class predictions as returned by NPClassifier of COCONUT compounds.</p> <p>Used NPClassifier between 2024.01.03~2024.01.15. Used COCONUT version 2022.01.01</p> <p> </p> <p> </p> <p>References:</p> <ol> <li>Kim, H. W. et al. NPClassifier: A Deep Neural Network-Based Structural Classification Tool for Natural Products. J. Nat. Prod. 84, 2795–2807 (2021). 10.1021/acs.jnatprod.1c00399</li> <li>Sorokina, M., Merseburger, P., Rajan, K., Yirik, M. A. & Steinbeck, C. COCONUT online: Collection of Open Natural Products database. J. Cheminformatics 13, 2 (2021). https://doi.org/10.1186/s13321-020-00478-9 <p> </p> </li> </ol>
Data to Support Predictive Models for Detrital Titanite Provenance with application to the Nanga Parbat syntaxial massif, western Himalaya."
<p>The files published here are metadata that are being used to support a manuscript currently (Mar, 2024) undergoing final reviews in Journal of Geophysical Research: Earth Surface.</p> <p>The intention of these data and code is to support a publication that is about generating a predictive categorisation scheme for the mineral titanite.</p> <p>The code to generate the titanite classification schemes was created in Python3, using Jupyter Notebook. The files also provide more motivation for why a predictive categorisation scheme for the mineral titanite is desirable, and other similar context. Chiefly, the dataset and random forest models published here will allow us to trace titanite in detritus.</p> <p>For info on running Jupyter Notebook, please visit (<a href="https://jupyter-notebook-beginner-guide.readthedocs.io/en/latest/execute.html">https://jupyter-notebook-beginner-guide.readthedocs.io/en/latest/execute.html</a>) to seek instructions. We also provide a readme file with some instructions. If you get really stuck, just email the authors.</p> <p>Our Model can be compared to similar previously published works (e.g. <a href="https://doi.org/10.1111/ter.12574">https://doi.org/10.1111/ter.12574</a>). Model was trained using skikit-learn v1.41.</p> <p>The supplementary file "Table_S4_Merged.csv" was used to train and generate the model.</p> <p>Your unknowns must contain the correct elements and labelling for the code to successfully run, these details are provided in the code (Titanite_Random_Forest_Model1_Mar24.ipynb). A template is also provided for you to paste your unknown data into (titanite_data_template.csv)</p> <p>Any new published data are titanite compositional or isotopic data collected by LA-ICP-MS. Description of how those data were collected is given in "OSullivan_et_al_Supp..." file.</p> <p>Some of the data, information and code in this submission has been subject to change after journal review, this is a second version of this content.</p> <p>References for the dataset compilation are provided in File S3.</p> <p>If you have any queries contact:<br>Gary O'Sullivan, Trinity College Dublin</p>
Predicted genes from the Amblyomma americanum draft genome assembly
<p>Data for pub "Predicted genes from the <em>Amblyomma americanum </em>draft genome assembly."</p> <ul> <li>Amblyomma_americanum_filtered_assembly.fasta: Decontaminated A. americanum genome with bacterial contigs removed</li> <li>Amblyomma_americanum_bacterial_contigs_info.tsv: Information about contigs classified as bacteria that were removed</li> <li>Amblyomma_americanum_annotation_data.tar.gz: Directory of annotation data produced by EvidenceModeler as part of the nf-core/genomeannotator workflow. Includes files in FASTA format (predicted genes and proteins), set of proteins clustered at 99% identity in FASTA format, and annotations in both GFF3 and GTF formats. GTF file produced from the GFF3 file with AGAT.</li> <li>Amblyomma_americanum_transcriptome_assembly_data.tar.gz: Directory of data generated for the transcriptome assembly that was used for gene prediction</li> </ul>
SkinSource: A Data-Driven Toolbox for Predicting Touch-Elicited Skin Vibrations Across the Upper Limb
<p>The repository contains the data for the toolbox released as part of the publication “SkinSource: A Data-Driven Toolbox for Predicting Touch-Elicited Vibrations in the Upper Limb.” The toolbox and installation and usage instructions can be found on GitHub here: <a href="https://github.com/neelitummala/skinsource">https://github.com/neelitummala/skinsource</a>. If you use these data or our toolbox please cite our publication: <a href="https://doi.org/10.1109/HAPTICS59260.2024.10520852">https://doi.org/10.1109/HAPTICS59260.2024.10520852</a>.</p> <p>Full citation: “Tummala, N., Reardon, G., Fani, S., Goetz, D., Bianchi, M., and Visell, Y. (2024) SkinSource: A Data-Driven Toolbox for Predicting Touch-Elicited Vibrations in the Upper Limb. IEEE Haptics Symposium 2024. DOI: 10.1109/HAPTICS59260.2024.10520852” </p> <p> </p> <p><strong>Abstract From Manuscript</strong></p> <p>Vibrations transmitted throughout the hand and arm during touch contact play a central role in haptic science and engineering but are challenging to model or experimentally characterize. Here, we present SkinSource, a data-driven toolbox for predicting skin vibrations across the upper limb in response to user-specified input forces. The toolbox leverages impulse response measurements that encode the physics of vibration transmission across the hands and arms of four participants and provides software tools for analyzing the predicted skin responses. We show that the SkinSource predictions closely match experimental measurements and confirm the underlying assumption of linear vibration transmission in the skin. We also demonstrate through several usage examples how SkinSource can act as a versatile computational platform for haptic research applications, such as characterizing vibrotactile transmission in the skin, engineering haptic interfaces, and investigating touch perception.</p> <p><strong> </strong></p> <p><strong>Dataset Description</strong></p> <p>This dataset comprises experimental data of 3-axis surface acceleration at 72 locations on the skin in response to unit impulsive forces supplied at 20 different input locations on the palmar hand surface. For details on our experimental procedure, please see our publication. This data is intended to be used as part of the SkinSource toolbox, which can be found here: <a href="https://github.com/neelitummala/skinsource">https://github.com/neelitummala/skinsource</a>.</p> <p><strong> </strong></p> <p><strong>Data Fields</strong></p> <p>The data is provided as a .mat file. This file contains a single variable “dataTable” of variable type “table.” The table contains 80 rows, each corresponding to a unique experimental condition (4 participants x 20 input locations), and contains the following fields:</p> <p><strong>Data </strong>(522x72x3) - 3D array containing the 3-axis skin acceleration at 522 time points (impulse responses) for each of 72 accelerometers. Please see the GitHub code and documentation (<a href="https://github.com/neelitummala/skinsource">https://github.com/neelitummala/skinsource</a>) for the accelerometer locations on the dorsal surface of the upper limb.</p> <p><strong>Model </strong>- The upper limb model number. This number specifies the participant that data was taken on.</p> <p><strong>Location</strong> -<strong> </strong>Number designating which input location on the palmar hand surface the data corresponds to. Please see the GitHub code and documentation (<a href="https://github.com/neelitummala/skinsource">https://github.com/neelitummala/skinsource</a>) for input location number mapping.</p>
Machine learning predicts earthquakes in the continuum model of a rate-and-state fault with frictional heterogeneities
<p>Numerical data used to make Figures in the manuscript entitled "Machine learning predicts earthquakes in the continuum model of a rate-and-state fault with frictional heterogeneities". We provide the data to create Figures 1 to 4 from the main text and Figures S1 to S9 from the supplementary information. We also provide Python scripts to plot them.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.