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 “predictions”
Aerial Images_Part 2_Integrating Remote Sensing and Machine Learning for Developing Spatio-Temporal Model to Predict Aquatic Larval Habitats of Malaria
<p>Aerial Images_Part 2_Integrating Remote Sensing and Machine Learning for Developing Spatio-Temporal Model to Predict Aquatic Larval Habitats of Malaria</p>
Aerial Images_Part 1_Integrating Remote Sensing and Machine Learning for Developing Spatio-Temporal Model to Predict Aquatic Larval Habitats of Malaria
<p>Aerial Images_Part 1_Integrating Remote Sensing and Machine Learning for Developing Spatio-Temporal Model to Predict Aquatic Larval Habitats of Malaria</p>
Supporting materials for: pKa Prediction in Non-Aqueous Solvents
<p>This repository includes datasets and supplementary materials for the manuscript: "pKa Prediction in Non-Aqueous Solvents" by Jonathan W. Zheng, Emad Al Ibrahim, and William H. Green. <strong>Citations should refer directly to the manuscript:</strong></p> <blockquote> <p>Zheng, J. W., Al Ibrahim, E., Kaljurand, I., Leito, I., & Green, W. H. (2024). pKa Prediction in Non-Aqueous Solvents. <em>Journal of Computational Chemistry (2024), </em>doi:10.1002/jcc.27517</p> </blockquote> <p>This compilation includes the predicted and experimental pKa values for all compounds studied in the corresponding work, as well as .xyz files corresponding to all conformers used in the COSMO-RS calculations. </p> <p>For the <strong>.csv </strong>files, column <code>pKa_exp</code> corresponds to the originally-reported experimental value whereas <code>pKa_OK</code> corresponds to the corrected value.</p> <p>The data from <strong>low_error_solvent_preds.csv</strong> and <strong>high_error_solvent_preds.csv</strong> and <strong>unreliable_solvent_preds.csv</strong> are formatted in part based on their compilation in the source manuscript: <em>Busch, M., Ahlberg, E., Ahlberg, E., & Laasonen, K. (2022). How to Predict the pKa of Any Compound in Any Solvent. ACS Omega, 7(20), 17369-17383.</em></p> <p> </p> <ul> <li>Note on version Aug. 28, 2024: fixed erroneous SMILES for several of the benzoic acids in the test set.</li> <li>Note on version Nov. 18, 2024: updated many of the values in the training and test sets, excluding some values and correcting others. SMILES (and corresponding .xyz files) for two values in the test set were updated. </li> </ul>
Electronic and optical properties of computationally Predicted Na-K-Sb crystals
<p>This is the dataset for the identically-named paper "Electronic and optical properties of computationally predicted Na-K-Sb crystals".</p> <p>This dataset contains output files from exciting for the two considered polymorphs: hexagonal Na2KSb (hNa2KSb.zip) and cubic NaK2Sb (cNaK2Sb.zip). Inside each zip file contains:</p> <ul> <li>INFO.OUT, INFOXS.OUT, GW_INFO.OUT, INFO_SCR.OUT</li> <li>Lattice-optimised input.xml files for both crystals</li> <li>BAND_Sxx_Ayyyy.OUT files for projected bandstructures</li> <li>PDOS_Sxx_Ayyyy.OUT files for projected density of states</li> <li>BAND.OUT, BANDLINES.OUT, BAND-QP.OUT, IDOS.OUT and TDOS/TDOS-QP.OUT for DOS</li> <li>EFERMI.OUT and EIGVAL.OUT</li> <li>EPSILON files from the dielectric tensor with/without the TDA and IQPA</li> <li>EXCITON files from the dielectric tensor with/without the TDA and IQPA</li> <li>KPATH files for considered excitons with the TDA</li> </ul>
Prioritization of unknown features based on predicted toxicity categories
<p> The fish toxicity data set, the filtered CompTox data set, and the pesticide mixture datasets used for the development of the prioritization algorithms described elsewhere.</p>
ND250 as a prediction error signal in orthographic processing: insights from the comparison of handwritten and printed words
<p><span>This dataset contains electroencephalography (EEG) recordings and behavioral data from a study investigating the neural mechanisms of visual word recognition in native Chinese speakers. The study used a color decision task, where participants viewed printed and handwritten Chinese single-character words varying in lexical frequency (high-frequency vs. low-frequency). The primary aim was to examine the N250 ERP component, a 250-ms difference in brain activity observed between certain word types, and determine whether it reflects activation of the orthographic lexicon or a prediction error signal during orthographic processing. The findings suggest that the N250 is related to prediction error, providing support for the Interactive Account of orthographic processing.</span></p>
Prediction of COVID-19 case numbers using state-space modeling and wastewater virus datasets
Open the record for dataset details and reuse information.
Processed Sentinel 1, Sentinel 2 and Copernicus Emergency Management Service data for fine tuning and predicting flood extent with IBM's granite-geospatial-uki-flood-detection model
<p>This dataset contains processed Sentinel 1 Sentinel 2 imagery together with flood event labels extracted from the Copernicus Emergency Management Service. It has been assembled to demonstrate fine tuning and inference of flood event segmentation using granite geospatial foundation models developed by IBM Research. Please see <a href="https://huggingface.co/ibm-granite/granite-geospatial-uki-flooddetection">https://huggingface.co/ibm-granite/granite-geospatial-uki-flooddetection</a> for more information on models and use.</p> <p>Sentinel-1</p> <p>The European Space Agency. 2014. Sentinel-1 Mission. <a href="https://sentinel.esa.int/web/sentinel/copernicus/sentinel-1">https://sentinel.esa.int/web/sentinel/missions/sentinel1</a>. Accessed: 2024-11-25.</p> <p>Sentinel-2</p> <p>The European Space Agency. 2015. Sentinel-2 Mission. <a href="https://sentinel.esa.int/web/sentinel/copernicus/sentinel-2">https://sentinel.esa.int/web/sentinel/missions/sentinel2</a>. Accessed: 2024-11-25.</p> <p>Copernicus Emergency Management Service</p> <p><a href="https://emergency.copernicus.eu/mapping/list-of-activations-rapid">https://emergency.copernicus.eu/mapping/list-of-activations-rapid</a>. Accessed: 2024-11-25. </p> <p><strong>Attribution</strong></p> <p>Contains modified Copernicus Sentinel data [2019-2024]</p> <p>Contains modified Copernicus Service information [2019-2023]</p>
Thermal coupling mode in mantle-outer core convection predicted from an ultra-high-resolution numerical simulation of two-layer convection with a large viscosity contrast
<p>Movie of temperature and velocity fields in the highly viscous layer (HVL) and the low-viscosity layer (LVL) (left panels) and the close-up views focusing on the interior of the LVL (right panels). The viscosity contrast between the HVL and LVL is 10<sup>4</sup>.</p>
Predicting transcriptional responses to novel chemical perturbations using deep generative model for drug discovery
<p>Understanding transcriptional responses to chemical perturbations is central to drug discovery, but exhaustive experimental screening of diseasecompound combinations is unfeasible. To overcome this limitation, here we introduce PRnet, a perturbation-conditioned deep generative model that predicts transcriptional responses to novel chemical perturbations that have never experimentally perturbed at bulk and single-cell levels. Evaluations indicate that PRnet outperforms alternative methods in predicting responses across novel compounds, pathways, and cell lines. PRnet enables gene-level response interpretation and in-silico drug screening for diseases based on gene signatures. PRnet further identifies and experimentally validates novel compound candidates against small cell lung cancer and colorectal cancer. Lastly, PRnet generates a large-scale integration atlas of perturbation profiles, covering 88 cell lines, 52 tissues, and various compound libraries. PRnet provides a robust and scalable candidate recommendation workflow and successfully recommends drug candidates for 233 diseases. Overall, PRnet is an effective and valuable tool for gene-based therapeutics screening.</p>
Machine learning prediction of enzyme optimal pH
<p>- Enzyme catalytic optimum pH dataset (pHopt, 9855 proteins)</p> <p>- Secreted bacterial optimum environment pH dataset (pHenv, 1.9 million proteins)</p> <p>- Model for predicting pHopt of enzymes (EpHod)</p> <p>- Code for using the EpHod model are in <a href="https://github.com/beckham-lab/EpHod">GitHub</a>. Paper in <a href="https://doi.org/10.1101/2023.06.22.544776">BioRxiv</a></p> <p> </p>
Tree census data associated to 'Large-scale informative priors to better predict the local occurrence rate of a rare tree-related microhabitat'
<p>This repository contains two excel spreadsheets providing data about the trees studied in Cottais et al. study (doi : 10.1101/2024.11.28.625900):</p> <ul> <li>df_TreM.xlsx has 1462 lines (headers not included); each line corresponds to one oak tree sampled in 2021 survey of Cottais et al. study. Fields are: <ul> <li>Plot_Id : the id of the sampling plot where the tree is located</li> <li>Tree_Id : a unique id of the tree used in the BloBiForM project</li> <li>Species_Latin_Name : the latin name of the tree species ('Quercus_sp' for all trees given than only oaks are reported here)</li> <li>DBH_cm : the diameter at breast height of the tree in cm</li> <li>TreM : whether the tree harbours a basal rot hole (1) or not (0)</li> <li>Life_Stage : whether the tree is a living tree or a snag</li> <li>Last_Logging_Year : the year of the last logging event in the stand where the tree is located</li> <li>conv : whether the last logging event occured less that 120 years before 2024 (conv=0; the stand is now classic high forest) or is more ancient (conv=1, the stand is being converted to high forest through sprout thinning)</li> </ul> </li> <li>DMH_2022.xlsx has 1190 lines (headers not included); each line corresponds to one tree sampled in 2022 survey of Cottais et al. study. Fields are: <ul> <li>Plot_Id: the id of the sampling plot where the tree is located</li> <li>Species_Latin_Name: the latin name of the tree species, obtained as a translation from french using the dictionnary in page 2; oaks are not the only species reported, but only oaks are used in Cottais et al. study;</li> <li>Species_French_Name : the french name of the tree species;</li> <li>DBH_cm: the diameter at breast height of the tree in cm</li> <li>TreM: whether the tree harbours a basal rot hole (1) or not (0)</li> <li>Tree_Id: when the tree belongs to the cohort of the BloBiForM project, the tree id is reported; left empty otherwise</li> <li>Nb_brin: number of stems on the tree stump;</li> <li>Commentaire: any comments during fieldwork</li> </ul> </li> </ul>
Dataset for "A Comparative Review of Deep Learning Methods for RNA Tertiary Structure Prediction"
<p>Datasets used in "A Comparative Review of Deep Learning Methods for RNA Tertiary Structure Prediction".</p> <p>The provided zip file contains:</p> <ul> <li><strong>Datasets 1–3:</strong> For each dataset, directories include input sequences (FASTA), multiple sequence alignments (MSAs in AFA format), and normalized predicted structures from six deep learning tools. Dataset 3 also contains references - RNA chains extracted from complexes. These folders also include a CSV file with all metrics for all RNAs and tools.</li> <li><strong>Dataset 4:</strong> A text file listing the PDB IDs of RNAs included in this dataset, which is a subset of Dataset 3.</li> <li><strong>Dataset complexes:</strong> RNA chains extracted from predicted complexes, where predictions are made by AlphaFold 3 web server and, in some cases, RoseTTAFoldNA. There are also job files for the AlphaFold 3 web server used to obtain these predictions. Same as for previous datasets, this folder also includes a CSV file with all metrics for these RNA chains.</li> </ul> <p>The structure of the dataset and details of each folder are explained in the included README file.</p> <p> </p>
Predicted structures of SLC-protein complexes and controls
<p>Solute carrier (SLC) transporters form a major superfamily which transport a wide range of substrates across cellular and organellar membranes. To study their regulation on protein level and position SLCs in the human interactome, we conducted a large-scale interrogation of the protein-protein interactions (PPIs) of SLCs employing affinity purification combined with mass spectrometry (AP-MS). This study resulted in thousands of novel protein interactions of SLCs. For a subset of SLC protein complexes, we performed structural predictions using AlphaFold (v2.2, v2.3 and v3). The dataset attached contains the structures in PDB/CIF format. The structures were further discussed in the associated manuscript. In addition, an annotation table is provided, which summarizes the scores for each modelled complex. </p>
CARA: Benchmarking Compound Activity Prediction for Real-World Drug Discovery Applications
<p>Identifying active compounds for target proteins is fundamental in early drug discovery. Recently, data-driven computational methods have demonstrated promising potential in predicting compound activities. However, there lacks a well-designed benchmark to comprehensively evaluate these methods from a practical perspective. To fill this gap, we propose a benchmark, named CARA.Through carefully distinguishing assay types, designing train-test splitting schemes and selecting evaluation metrics, CARA can consider the biased distribution of current real-world compound activity data and avoid overestimation of model performances. We observed that current models can make successful predictions for certain proportions of assays, while the performances varied across different assays. In addition, evaluation of several few-shot training strategies demonstrated different performances related to task types. Overall, we provide a high-quality dataset for developing and evaluating compound activity prediction models, and the analyses in this work may inspire better applications of data-driven models in drug discovery.</p>
Predictive Maintenance Literature Review Framework
<p>This dataset contains information collected during a literature review in the domain of Predictive Maintenance in Industry 4.0. It consists of Two Tables:</p><ol><li>Paper references and their contribution to individual topics</li><li>Types of Neural Networks that have been used in the Papers</li></ol>
Proteome-wide Prediction of the Functional Impact of Missense Variants with ProteoCast
<p><strong>Proteome-wide Prediction of the Functional Impact of Missense Variants with ProteoCast</strong></p> <p>This dataset contains mutation effect predictions for 22,169 <em>Drosophila melanogaster</em> protein isoforms, classifying over <strong>293 million </strong>amino acid substitutions as neutral, uncertain, or impactful. The predictions were generated using the evolution-based <strong>GEMME </strong>model (E.Laine et <em>al.</em> MBE 2019) with multiple sequence alignments (MSAs) from the highly efficient <strong>ColabFold</strong> protocol (M.Mirdita et<em> al.</em> NatMet 2022, Abakarova et al. <em>GBE</em> 2023). To ensure reliability, we provide<strong> </strong>global (per-protein) and local (per-residue) confidence metrics, since the predictions are sensitive to the input MSA quality.</p> <p>Predictions were validated using natural polymorphisms from the Drosophila Genetic Reference Panel (DGRP) and Drosophila Evolution over Space and Time (DEST2) datasets, as well as FlyBase’s developmentally lethal and hypomorphic mutations. Additionally, the dataset includes sensitivity data for post-translational modifications (PTMs) and short linear motifs (SLiMs), aiding functional site identification.</p> <p>All this data can be visualized at <a href="https://proteocast.ijm.fr/drosophiladb/">proteocast.ijm.fr</a>. </p> <p>Readme:</p> <ol> <li><strong>Drosophila_ProteoCast.tar.gz </strong>- this archive contains ProteoCast predictions and analysis for each unique proteoform, with folder names corresponding to IDs listed in the mapping_database.csv file. A detailed description of the folder structure and contents can be found in ReadMe.txt. </li> <li><strong>data.tar.gz </strong>- this archive contains the data used in this study, sourced from FlyBase, DGRP2, and DEST2.</li> <li><strong>Dmel6.44PredictionsRecap.csv</strong> - this summary file provides detailed information for each proteoform, all FlyBase protein IDs (<em>FBpp_ID</em>) included. It contains the following data: <ul> <li><strong>Identifiers</strong>: FlyBase protein ID (<em>FBpp_ID</em>), protein symbol (Protein_symbol), gene ID (<em>FBgn_ID</em>), transcript ID (<em>FBtr_ID</em>), and UniProt ID if available (<em>UniProt_ID</em>).</li> <li><strong>Protein Characteristics</strong>: Sequence length (<em>Length</em>) and whether the proteoform is representative (<em>Representative_FBpp</em>).</li> <li><strong>MSA and GEMME Predictions</strong>: Fraction of observed mutations (<em>F_obs</em>) and number of sequences (<em>Nb_seq_MSA</em>) in the ColabFold MSA, presence or absence of GEMME predictions (<em>GEMME_prediction</em>), and global confidence score (<em>GlobalConfidence</em>).</li> <li><strong>Mutation Classification</strong>: Thresholds for defining mutations as neutral, uncertain, or impactful (<em>GMM3_uncertain</em>, <em>GMM3_impactful</em>).</li> <li><strong>Genomic Information</strong>: DNA strand (<em>Strand</em>) and exon coordinates (<em>Exons_coordinates</em>).</li> <li><strong>Structural Data</strong>: 3D structure file name if available (<em>Structure_3D_file, Structure_3D</em>).</li> <li><strong>Mutation Counts</strong>: Number of analyzed mutations and affected residues, labelled as lethal, hypomorphic on FlyBase, or from the DEST2 and DGRP datasets (<em>n_Lethal, n_Lethal_res, n_Hypomorphic, n_Hypomorphic_res, n_DEST2, n_DEST2_res, n_DGRP, n_DGRP_res, n_DEST_DGRP_union, n_DEST_DGRP_union_res</em>).</li> </ul> </li> <li><strong>csv.tar.gz </strong>-<strong> </strong>this archive contains the files generated in this study. A detailed description of the folder structure and contents can be found in ReadMe.txt.</li> </ol>
Geometric deep learning improves generalizability of MHC-bound peptide predictions
<p>Full dataset and trained models from the manuscript "<strong>Geometric deep learning improves generalizability of MHC-bound peptide predictions</strong>".</p> <p>"outputs_and-BA_data.zip" contains the networks' outputs for each cross-validation experiment and a "full_dataset.csv" containing the initial BA data.<br>Note: this file has been updated (2024/11/26) due to errors in generating some of the previous csvs. In the earlier version, both MLP and CNN outputs reported were wrong. The correct values are now reported in the updated csvs.</p> <p>"trained_models.zip" contains all the trained models parameters</p> <p>"propedia_ssl.zip" contains all the 3D models from propedia used to train the 3D-SSL</p> <p>"pdb.zip" contains 3D models generated in PANDORA and used to train CNN, GNN and EGNN. It amounts to 145665 .pdb files, one for each human binding affinity entry from the initial dataset from O'Donnell et al. The list of entries used to actually train networks after filtering can be found in outputs_and-BA_data.zip", in the "full_dataset.csv" file. </p> <p> </p> <p>CHANGELOG v4:</p> <p>- In outputs_and-BA-data.zip, updated CNN_AlleleClustered_test_crossval.csv and CNN_shuffled_test_crossval.csv. These file had the wrong IDs paired with the network outputs.The IDs and labels are now consistent with the outputs.</p> <p>- Updated reference from the preprint to the published article. </p> <p> </p>
Machine Learning for predicting chaotic systems – Data
<p>The data used in our article "Machine Learning for Predicting Chaotic Systems" - <a href="https://arxiv.org/abs/2407.20158">https://arxiv.org/abs/2407.20158</a></p> <p>DeebDbDysts*.zip contain the Dysts database, DeebDbLorenz*.zip the DeebLorenz database (with DeebDbLorenzBig*.zip being the "extension" dataset for Lorenz63std with different time series lengths).</p> <p>The observation and truth data of the Dysts database originates from <a href="https://github.com/williamgilpin/dysts">https://github.com/williamgilpin/dysts</a> (we converted the data format from json to csv).</p> <p>For DeebLorenz, we used the R package <a href="https://github.com/chroetz/DEEBdata">DEEBdata</a> to create it.</p>
PRESCOTT/ESCOTT/iGEMME mutational effect predictions of all single point mutations for ~3000 proteins
<p>This dataset contains all necessary data to reproduce our analyses on ~3000 proteins:</p> <p>It is made up of 4 compressed datasets. </p> <ol> <li>All colabfold MSAs and structures for ~3000 proteins: colabfold-sequences-structures-3000-proteins.tar.bz2</li> <li>All escott prediction: escott-v-1-6-0-max-two-components-colabfold-msas-entire-single-point-mutations-cvRC7.tgz</li> <li>All igemme predictions: escott-v-1-6-0-tjet-only-colabfold-msas-entire-single-point-mutations.tgz</li> <li>All gnomad v4.0.0 csv files used for prescott predictions: gnomadv4-0-0-csv-files.tgz</li> </ol> <p>The dataset contains list of 500 proteins used in determining PRESCOTT coefficients in TRAINING-all_gene_names_v4_set1_no_acmg.txt. </p> <p>Furthermore, the list of 1883 proteins used to measure for testing purpose is given in TESTING-all_gene_names_v4_set1_with_acmg.txt. </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.