Find research datasets worth reusing

opencc-zeroOct 2021View details →

B3DB Dataset adapted for A Transparent Machine Learning Model To Understand Drugs Permeability Through the Blood Brain Barrier

The B3DB dataset which we adapted for use for our paper "A Transparent Machine Learning Model To  Understand Drugs Permeability Through the Blood Brain Barrier"

Yields, Cannabinoids Quantification, and Predictive Programming Codes Using Machine Learning for Non-Psychoactive Cannabis Flowers and Extracts (Cannabis sativa L.) Cultivated in Ecuador.

This publication presents data from various extraction methods, including maceration, Soxhlet, and supercritical fluids, performed on different cannabis flower varieties (Cannabis sativa L.) under varying operating conditions. We quantified the amounts of CBD, THC, CBG, and CBN in the extracts produced by each method using High-Performance Liquid Chromatography (HPLC). Using this data, we developed a machine learning algorithm in RStudio to make predictions and determine the best conditions and yields for each extraction method. The analysis focuses on different varieties of non-psychoactive cannabis cultivated in Ecuador at altitudes over 2,450 m.a.s.l.

Mapping Soil Organic Carbon in the World's Largest Arid Mangrove Forest (Indus Delta, Pakistan): A Multi-Sensor Remote Sensing and Machine Learning Approach

Mangrove forests play a crucial role in carbon sequestration, especially in arid regions where their ability to store carbon in soil is vital for mitigating climate change. The Indus Delta in Pakistan, the world’s largest arid mangrove forest system, lacks spatially explicit data on Soil Organic Carbon (SOC) despite its importance for conservation and carbon budgeting. This study aims to establish a baseline SOC map 2020 at 10 m spatial resolution using Sentinel-1 (Synthetic Aperture Radar) and Sentinel-2 (MultiSpectral Instrument) satellite imagery, integrated with in-situ soil sampling. SOC predictions were made using a Classification and Regression Tree (CART) machine learning model within the Google Earth Engine platform, leveraging 40 predictor variables, including spectral bands and derived indices. A total of 53 topsoil (0-10 cm) samples were collected in February 2020 across the Indus Delta, and SOC was analyzed using the Walkley-Black method. The results showed an average SOC value of 65.88 Mg C ha⁻¹ with substantial spatial variability, ranging from 15.06 Mg C ha⁻¹ to 138.03 Mg C ha⁻¹ with a total of 0.91 Pg C. The CART model demonstrated high accuracy, with an R² of 0.95 and an RMSE of 9.18 Mg C ha⁻¹. However, the region faces challenges such as seawater intrusion and salinity, which threaten its ability to sequester carbon. With the first high-resolution SOC map for the Indus Delta, this study provides valuable insights for ecosystem management, conservation planning, and carbon budgeting. These findings of this study have the potential to significantly influence initiatives like REDD+ and Blue Carbon projects, which aim to enhance carbon sequestration while addressing the ecological challenges facing Pakistan’s mangroves

Impact of Interval Censoring on Data Accuracy and Machine Learning Performance in Biological High-Throughput Screening

<div> <h2>Overview</h2> <div>Data and Results used in the publication entitled "Impact of Interval Censoring on Data Accuracy and Machine Learning Performance in Biological High-Throughput Screening"</div> </div> <div> <h3>Data</h3> <div>This folder contains the raw data used during this work.</div> <div>`EvoEF.csv` contains information on the library used (sequences, number of mutations, etc.) and the fitness (energy) used as continuous mean values. `mut.csv` contains the information about the combinatorial scaling (N vs N_norm), the number of mutations (m) and the probability of each variant using different distributions (uniform and binomial) at different $p_{WT}$.</div> <div>For further details on how the fitness values were calculated and how the combinatorial scale works, please refer to our prevoius [Paper](https://arxiv.org/abs/2405.05167).</div> <div> </div> <h3>Results</h3> <div> <div>This folder contains the results (outputs) of all scripts used. Such results are included in the form of `.npy` and `.npz` files. To load such files with numpy you should include the option `allow_pickle=True`.</div> </div> </div>

KuafuPrimer: Machine learning facilitates the design of 16S rRNA gene primers with minimal bias in bacterial communities

KuafuPrimer is a machine learning-aided method that learns community characteristics from several samples to design 16S rRNA gene primers with minimal bias for microbial communities. It is built on Python 3.9.0, Pytorch 1.12.0. Here are some large size files required to run KuafuPrimer, and users need to download and put them in correct directories before running the program. <ol> <li>Silva_ref_data.zip: processed files of silva dataset that should be put in <code>Model_data/Silva_ref_data/</code>.</li> <li>DeepAnno16_publicated_model.zip: parameters of the trained DeepAnno16 model that should be put in <code>Model_data/DeepAnno16_publicated_model/</code> .</li> </ol> For more information, please refer to https://github.com/zhanghaoyu9931/KuafuPrimer.

Research data for "Exploring the energy landscape of aluminas through machine learning interatomic potential"

This dataset supports the paper "Exploring the energy landscape of aluminas through machine learning interatomic potential". The paper is online here: The following folders are provided: <ul> <li>classical_potential_files: Contains all the empirical potentials used in this study.</li> </ul>   <ul> <li>crystal_structure_file: Contains structural files of aluminas with various crystal structures, which can be distinguished by their respective filenames. Configurations of alumina with partially occupied cation sites can be obtained from the references provided in the supplementary materials of our article.</li> </ul>   <ul> <li>lowest_E-config: The files named <code>poscar_{0..19}</code> represent the 20 structure files identified through our developed structural search workflow in conjunction with the final NEP of Aluminas. These structures exhibit different distributions of Al cation occupancy sites. The suffix numbers in the file names indicate that these 20 structures are arranged in ascending order based on their corresponding energy values after structural relaxation using the NEP. In other words, <code>poscar_0</code>, after structural optimization, has the lowest energy among these 20 configurations. Additionally, we have included the CIF files for the crystal structures with partial occupancies provided by the Smrcok model in this folder.</li> </ul>   <ul> <li>the_final-dataset_alumina: This folder contains all the relevant files for training the final NEP of Aluminas, including the final training dataset named <code>train.xyz</code>, the training parameter file <code>nep.in</code>, and log files. The file <code>nep.txt</code> refers to the final NEP of Aluminas. </li> </ul>   <ul> <li>various_test-datasets: This folder provides all the test datasets used for testing the final NEP of Aluminas. We have categorized them into four types based on composition: clusters, amorphous structures, crystals, and datasets with physically unallowed configurations that exhibit nearest-neighbor cation occupancy according to the Smrcok model.</li> </ul> Additionally, for ease of retrieval, we have placed the file for the final NEP of aluminas in the main directory and named it <code>nep_3335.txt</code>, where the suffix indicates that the final training dataset <code>train.xyz</code> contains 3,335 structures. This file is identical to the file named <code>nep.txt</code> located in the folder <code>the_final-dataset_alumina</code>.

p-IgGen Dataset: Cleaned paired and unpaired antibody sequence data for machine learning applications.

This data is released alongside "p-IgGen: A Paired Antibody Generative Language Model", which contains full details on the data processing and cleaning. p-IgGen Paper: https://www.biorxiv.org/content/10.1101/2024.08.06.606780v1 . OAS: https://opig.stats.ox.ac.uk/webapps/oas/

Datasets for "Accurate and efficient structure elucidation from routine one-dimensional NMR spectra using multitask machine learning"

This upload contains the datasets used for the experiments in: Accurate and efficient structure elucidation from routine one-dimensional NMR spectra using multitask machine learning Frank Hu, Michael S. Chen, Grant M. Rotskoff, Matthew W. Kanan, and Thomas E. Markland https://arxiv.org/abs/2408.08284   For file descriptions and usage, please refer to the supplied README.md file.

Machine learning reveals the diversity of human 3D chromatin contact patterns (example predictions genome wide)

Example data for the paper: Machine learning reveals the diversity of human 3D chromatin contact patterns GitHub: https://github.com/erin-n-gilbertson/3DGenome-diversity/tree/main biorXiv: https://www.biorxiv.org/content/10.1101/2023.12.22.573104v1.full Manuscript accepted at Molecular Biology and Evolution Of primary interest will be the example predictions genome wide for hg38 reference, human-archaic hominin ancestor and most divergent 1KG individual per genome along with the Jupyter notebook tutorial for making your own Akita predictions given any input 1MB sequence. <div> <ul> <li>bin: contains python script for and qsub array shell script for generating example predictions. These scripts can be modified to take in any fasta files as input.</li> <li>akita_predictions: contains both Akita prediction output arrays and SVG files with predicted contact maps for the hg38 reference, human-archaic hominin ancestor and most divergent 1KG individual in each of 4,873 1MB windows</li> <li>anc_window_spearman.csv: spearman correlation between each 1KG individual and the ancestor for each 1MB window. To calculate 3D divergence subtract these values from 1.</li> <li>basenji: basenji dir from their github, necessary in the directory to run predictions - https://github.com/calico/basenji/tree/master</li> <li>genomes: fasta genomes for hg38 reference and human-archaic hominin ancestor used to make akita predictions</li> <li>divergent_windows: variants and expected divergence distributions for 392 more divergent than expected windows. Defined in the manuscript as windows where 3D divergence between 1KG indiivudals and the ancestor is greater than what would be expected based on sequence divergence. See manuscript Fig. S9 for more details. </li> <li>windows.txt: 4,873 1MB genomic windows with 100% coverage in hg38 used for Akita predictions</li> <li>making_examples.ipynb: jupyter notebook with tutorial instructions for making Akita predictions on any human genome sequence.</li> </ul> </div>

Supplementary files: Machine Learning Insights into Türkiye's Climate Variability: Predictive Modelling and Spatial Analysis

This dataset and python code were used in the study titled "Machine Learning Insights into Türkiye's Climate Variability: Predictive Modelling and Spatial Analysis".

Reproducibility Case Study and Survey: Machine Learning-based Additive Manufacturing Process Monitoring and Quality Prediction

Machine learning (ML)-based monitoring systems have been extensively developed to enhance the print quality of additive manufacturing (AM). However, the reproducibility of the proposed ML-based AM monitoring systems in published works has not been investigated due to a lack of evaluation methods. In the paper 'Towards reproducible machine learning-based process monitoring and quality prediction research for additive manufacturing,' we propose a reproducibility investigation pipeline and conduct two case studies to validate the pipeline. This dataset records the data generated by one of the case studies. This dataset also contains the reproducibility survey results.

Automated Model Discovery for Tensional Homeostasis: Constitutive Machine Learning in Growth and Remodeling (Source code and data)

This dataset contains <ul> <li>the source code</li> <li>the data and examples</li> <li>the material subroutine with examples of uniaxial strain and stress</li> </ul> of the inelastic Constitutive Artificial Neural Network (iCANN) enhanced by the concept of homeostatic surfaces to discover tensional homeostasis. The corresponding publication is: Holthusen, H., Brepols, T., Linka, K., & Kuhl, E.. Automated Model Discovery for Tensional Homeostasis: Constitutive Machine Learning in Growth and Remodeling.   Standalone_Materialroutine <ul> <li>00_Materialroutine: Contains the material subroutine implemented in FORTRAN</li> <li>01_uniaxial_strain: Example of the material subroutine in a uniaxial strain driven manner</li> <li>02_uniaxial_stress: Example of the material subroutine in a uniaxial stress driven manner</li> </ul>   TensorFlow <ul> <li> iCANN: <ul> <li> 01_Biax/biax_l1: Keras/TensorFlow implementation of the iCANN. Example of the cross specimen with L1 (Lasso) regularization </li> <li> 01_Biax/biax_l2: Keras/TensorFlow implementation of the iCANN. Example of the cross specimen with L2 (ridge) regularization </li> <li> 02_Uniax/uniax_l1: Keras/TensorFlow implementation of the iCANN. Example of the stripe specimen with L1 (Lasso) regularization </li> <li>02_Uniax/uniax_l1: Keras/TensorFlow implementation of the iCANN. Example of the stripe specimen with L2 (ridge) regularization</li> </ul> </li> <li> iCANN_ABS_activation: Same four examples as above, however, with the absolute value activation function </li> <li> installed_packages: File containing a list of installed Python modules used to implement the iCANN </li> </ul> The TensorFlow implementations in all 01_Biax/ and 02_Uniax/ sub-directories are the same. The implementation in iCANN_ABS_activation is different with respect to the activation functions of the pseudo potential.   The experimental data for the cross and stripe specimen are taken from the literature: Eichinger, J. F., Paukner, D., Szafron, J. M., Aydin, R. C., Humphrey, J. D., & Cyron, C. J. (2020). Computer-controlled biaxial bioreactor for investigating cell-mediated homeostasis in tissue equivalents. Journal of biomechanical engineering, 142(7), 071011. <a href="https://doi.org/10.1115/1.4046201">https://doi.org/10.1115/1.4046201</a>

opencc-by-4.0Jun 2024View details →

Additional files for Horvath et al., 2024. Detection and classification of long terminal repeat sequences in plant LTR-retrotransposons and their analysis using explainable machine learning.

Additional data for Horvath et al., 2024 (source code freeze, models, data, supplementary figures, tables and files(.

opencc-by-4.0Feb 2025View details →

Geospatial micro-estimates of slum populations in 129 Global South countries using machine learning and public data

Reliable estimation of populations living in slums or slum-like conditions is crucial for urban planning, humanitarian resource allocation, and human well-being improvement. We generate the micro-estimate of slum population at a neighborhood level (~3.63 arc-minutes, preserving the privacy of vulnerable people) for 129 Global South countries in 2018. The estimates are built based on the Sustainable Development Goals 11.1 indicator framework and machine learning algorithms to heterogeneous data from household-based surveys and satellite images, as well as grided population data. Our integrated regional models show strong predictive capabilities for cluster-level slums proxy, explaining 82% to 96% of the variation in ground-truth surveys conducted in Global South countries, with root mean squared error ranging from 4.85% to 10.47%. The models perform match or surpass benchmarks established by previous studies. Cross-comparison with independent data sources at multi-scales suggest that our approach can yield reliable and consistent slum population estimates.

Reproduction package for: Using machine learning techniques to mitigate confidentiality violations

This is a reproduction package for the data shown in my master's thesis "Using machine learning techniques to mitigate confidentiality violations"

ALPHA-g Preprocessed Simulation Dataset for Machine Learning

Dataset generated for "Vertex Reconstruction with Deep Learning for ALPHA-g Radial Time Projection Chamber" 2024 preprint and "AI Meets Antimatter: Unveiling Antihydrogen Annihilations" paper for Machine Learning and the Physical Sciences workshop at NeurIPS'24 by Ferreira et al. The goal of this project was to create a PointNet-like deep learning model that could learn the relationship between simulated ALPHA-g detector data and the position of an antimatter annihilation on the walls of the ALPHA-g vacuum chamber. If this approach to antimatter annihilation event reconstruction works with real-world data then it can then be used as part of a larger analysis pipeline to better constrain the effect of gravity on antimatter. See the above papers for a much more detailed description of the experiment and data used but generally, the four important terms to know for interacting with this dataset are: <table> <tbody> <tr> <td>Term</td> <td>Meaning</td> </tr> <tr> <td>Event</td> <td>An antimatter annihilation. Each event contains a number of spacepoints and one vertex.</td> </tr> <tr> <td>Spacepoints</td> <td>The many 3D points (x,y,z) that represent the ALPHA-g detector data, if more than 800 this dataset cuts them off at 800, prioritizing the ones with higher amplitudes, and if less than 800 it pads them with zeros such that all events have 800 entries.</td> </tr> <tr> <td>Vertex</td> <td>3D position of the antimatter annihilation. For this initial study, we just use the z coordinate since this is what is most important to measuring the effect of gravity. </td> </tr> <tr> <td>Helix Fit</td> <td>The state-of-the-art method currently used for antimatter annihilation event reconstruction within the ALPHA-g detector. The method doesn't use machine learning and instead involves fitting 3D helix functions to the spacepoints.</td> </tr> </tbody> </table>   Data is split into three separate HDF5 datasets: training (2,160,020 events), validation (261,654 events), and testing (288,919 events). Each dataset contains the following columns with each row representing one event: <table> <tbody> <tr> <td>Column Index</td> <td>Value [mm]</td> </tr> <tr> <td>0</td> <td>x coordinates of spacepoints</td> </tr> <tr> <td>1</td> <td>y coordinates of spacepoints</td> </tr> <tr> <td>2</td> <td>z coordinates of spacepoints</td> </tr> <tr> <td>3</td> <td> z coordinate of simulated vertex </td> </tr> <tr> <td>4</td> <td> z coordinate of Helix Fit vertex prediction </td> </tr> </tbody> </table>   The code used to generate the raw dataset is available at <code><a href="bitbucket.org/expalpha/alphasoft">bitbucket.org/expalpha/alphasoft</a></code> and the code used to process it into this preprocessed dataset as well as the downstream training and evaluation are available at <code><a href="gitlab.triumf.ca/alpha-ai/rTPC-AI">gitlab.triumf.ca/alpha-ai/rTPC-AI</a></code>.

Machine Learning for Energy Consumption Prediction of Numerical Controlled Programs - NC Files

The NC files housed within this DOI represent the Data the Machine Learning Models were Trained/Validated/Tested on, during the execution of the work in the thesis, " Machine Learning for Energy Consumption Prediction of Numerically Controlled Programs." Theses files were created by Samuel D. Stencel, a Gradute Research Assistant at Purdue University.

opencc-by-4.0Jul 2024View details →

Data for "Machine learning of reduced quantum channels on NISQ devices"

This dataset contains the data, figures and code of the publication <a href="https://doi.org/10.48550/arXiv.2405.12598">"Machine learning of reduced quantum channels on NISQ devices"</a>. I.e., LeNoM (Learning Noise Models) represents the core implementation of our approach.

opencc-by-4.0Nov 2024View details →

Machine-learning-based validation of Microsoft Azure Kinect in measuring gait profiles

The dataset was used to validate Microsoft Azure Kinect in measuring gait profiles, employing machine learning techniques to investigate the impact of residual errors due to environmental, methodological, and processing factors on the accuracy of gait profile assessments. Data were collected from healthy and post-stroke subjects using a motion capture system and a 3D camera-based system with MAK, and corresponding gait profiles were estimated and compiled into a dataset. The estimated gait profiles include spatiotemporal, asymmetry, and body center of mass parameters to capture various normal and pathological gait characteristics. Contact e-mail: claudia.ferraris@cnr.it (Claudia Ferraris) lucavisma@hotmail.com (Luca Vismara) veronica.cimolin@polimi.it (Veronica Cimolin)