Find research datasets worth reusing

Dataset for: Exploring Battery Cathode Materials in the Li-Ni-O Phase Diagrams using Structure Prediction

The Li-Ni-O phase diagram contains several electrochemically active ternary phases. Many compositions and structures in this phase space can easily be altered by (electro-)chemical processes, yielding many more (meta-)stable structures with interesting properties. In this study, we use ab initio random structure searching (AIRSS) to accelerate materials discovery of the Li-Ni-O phase space. We demonstrate that AIRSS can efficiently explore structures (e.g. LiNiO2) displaying dynamic Jahn-Teller effects. A thermodynamically stable Li2Ni2O3 phase which reduces the thermodynamic stability window of LiNiO2 was discovered. AIRSS also encountered many dynamically stable structures close to the convex hull. Therefore, we confirm the presence of metastable Li-Ni-O phases by revealing their structures and properties. This work will allow Li-Ni-O phases to be more easily identified in future experiments and help to combat the challenges in synthesizing Li-Ni-O phases. This dataset contains the raw research data and key analysis files for "Exploring Battery Cathode Materials in the Li-Ni-O Phase Diagrams using Structure Prediction".  <ul> <li>`known_phases_aiida_data.zip` and `new_phases_exports.aiida.zip` contain the archives exported from the <a href="https://www.aiida.net">AiiDA framework</a> which was used to perform parts of the DFT calculations for this project. </li> <li>`search_data.zip` contains search seed files and the structures generated by the searches.</li> <li>`data_analysis.zip` contains the data and notebooks to reproduce the figures and tables shown in the manuscript.</li> </ul>

opencc-zeroJun 2023View details →

Reference data and simulated communities for 16S rRNA GCN prediction

16S rRNA gene has been widely used in microbial diversity studies to determine the community composition and structure. 16S rRNA gene copy number (16S GCN) varies among microbial species and this variation introduces biases to the relative cell abundance estimated using 16S rRNA read counts. To correct the biases, methods (e.g., PICRUST2) have been developed to predict 16S GCN. 16S GCN predictions come with inherent uncertainty, which is often ignored in the downstream analyses. However, a recent study suggests that the uncertainty can be so great that copy number correction is not justified in practice. Despite the significant implications in 16S rRNA-based microbial diversity studies, the uncertainty associated with 16S GCN predictions has not been well characterized. Here we develop a novel method to better model and capture the inherent uncertainty. Using cross-validation, we show that our method provides robust confidence estimates for the GCN predictions and outperforms PICRUST2 in both precision and recall. We found that 16S GCN correction should improve compositional and functional profiles estimated using 16S rRNA reads. On the other hand, we found that GCN variation has limited impacts on PCoA, PERMANOVA and random forest test, and 16S rRNA GCN correction is unnecessary in beta-diversity analyses.

opencc-zeroJun 2023View details →

Data and code from: Dynamics of cortical contrast adaptation predict perception of signals in noise

Neurons throughout the sensory pathway adapt their responses depending on the statistical structure of the sensory environment. Contrast gain control is a form of adaptation in the auditory cortex, but it is unclear whether the dynamics of gain control reflect efficient adaptation, and whether they shape behavioral perception. Here, we trained mice to detect a target presented in background noise shortly after a change in the contrast of the background. The observed changes in cortical gain and behavioral detection followed the dynamics of a normative model of efficient contrast gain control; specifically, target detection and sensitivity improved slowly in low contrast, but degraded rapidly in high contrast. Auditory cortex was required for this task, and cortical responses were not only similarly affected by contrast but predicted variability in behavioral performance. Combined, our results demonstrate that dynamic gain adaptation supports efficient coding in auditory cortex and predicts the perception of sounds in noise.

Data for internet traffic prediction using RNN

Contains three .csv files which are from Wireshark captures. The device used for capturing was a MacBook Pro 2019: - Game streaming data captured using PS Remote Play. - Video streaming data captured from a movie in Prime Video. - File download data captured from downloading a 6GB file from Google Drive.

Predicting Mathematics Anxiety and Achievement: Unveiling the Significance of Student and Teacher Attributes through Hierarchical Linear Modeling

This study aimed to determine the predictive power of student and teacher characteristics on students' math anxiety and achievement.

opencc-by-4.0Apr 2023View details →

Data sets and machine learning models for: Predicting critical properties and acentric factor of fluids using multi-task machine learning

The experimental data sets, data splits, additional features, QM calculations, model predictions, and final machine learning models for the manuscript "Predicting Critical Properties and Acentric Factor of Fluids Using Multi-Task Machine Learning". Citation should refer directly to the manuscript: <ul> <li> Biswas, S.; Chung, Y.; Ramirez, J.; Wu, H.; Green, W. H. Predicting Critical Properties and Acentric Factors of Fluids Using Multitask Machine Learning. Journal of Chemical Information and Modeling. 2023 63 (15), 4574-4588. DOI: <a href="https://doi.org/10.1021/acs.jcim.3c00546">10.1021/acs.jcim.3c00546</a> </li> </ul> To use the machine learning models, please refer to the sample files and instructions on <a href="https://github.com/yunsiechung/chemprop/tree/crit_prop">https://github.com/yunsiechung/chemprop/tree/crit_prop</a>.  Detailed information can be found in README.md file.   Details on the properties considered The data set includes the following 8 properties: <ul> <li>Tc: critical temperature, in K</li> <li>Pc: critical pressure, in bar</li> <li>rhoc: critical density, in mol/L</li> <li>omega: acentric factor, unitless</li> <li>Tb: boiling point, in K</li> <li>Tm: melting point, in K</li> <li>dHvap: enthalpy of vaporization at boiling point, in kJ/mol</li> <li>dHfus: enthalpy of fusion at melting point, in kJ/mol</li> </ul> Details on the files 1. Data sets under CritProp_v1.1.0: <ul> <li>all_data: includes the data sets used in this work. All data points are listed for each chemical compound as well as its corresponding data source. The details of the data sources can be found in the README.md file. The distribution of the data set is included in each folder. <ul> <li>estimated_data_for_pretraining: contains the estimated data from Yaws' handbook that are used to pre-train our machine learning (ML) model.</li> <li>experimental_data: contains the experimental data (references 1 - 15) used to fine-tune our final ML model.</li> </ul> </li> <li>additional_features: includes the additional features tested for the ML model. The Abraham features are generated for all data (references 1 - 15) while the acsf, qm, and rdkit features are only generated for the data from references 1 - 9. <ul> <li>abraham: Abraham solute parameters (E, S, A, B, L). Molecular features.</li> <li>acsf: ACSF (atom-centered symmetry functions). Atomic features that are coverted from the 3D coordinates of the compound</li> <li>qm_atom: QM (quantum chemical) atomic feature. </li> <li>qm_mol: QM molecular feature.</li> <li>rdkit: Selected RDKit 2D molecular features.</li> </ul> </li> <li>data_splits_and_model_predictions: contains the training and test sets used to evaluate the model. It also contains the predicted values from our final ML model for each test set. <ul> <li>random and scaffold splits: training and test sets that include the data from references 1 - 9.</li> <li>external test set: a test set that includes the data from only references 10 - 15.</li> </ul> </li> </ul> 2. Machine learning (ML) model files: <ul> <li>CritProp_ML_model_files_with_abraham_feat.zip: contains the Chemprop ML model files that are trained using Abraham features as additional molecular features. This gives the best results.</li> <li>CritProp_ML_model_files_without_additional_feat.zip: contains the Chemprop ML model files that are trained without any additional features. This gives the second best results.</li> </ul> To use these ML models, please refer to the sample files and instructions on <a href="https://github.com/yunsiechung/chemprop/tree/crit_prop">https://github.com/yunsiechung/chemprop/tree/crit_prop</a> 3. QM (quantum chemical) calculations: <ul> <li>QM_calculations.zip: contains the results of the QM calculations that are performed to compute QM features.</li> </ul>

Monitoring of postpartum body condition at the cow and herd levels: assessing explanatory and predictive power of disease risk models

Objectives 1- To define the herd threshold for cows with poor body condition based on its predictive capacity for disease risk at the herd level, and 2- to estimate the impact measures on disease rates due to body condition indicators in transition period. Two commercial grazing dairy herds (Herd A=5.034 and herd B=7.965 lactations) from Argentinean Pampa region were used to perform a longitudinal retrospective study during a 4-year period (2014 –2017).Health, reproductive and body condition score (BCS) records were gathered. The BCS (5-point scale) was performed at calving and at the time of reproductive release. The difference between both measures of BCS was used to assess the body condition loss (∆BCS). All the cows not bred by 70 DIM were checked for anestrus.Calving cohorts of 21-day were defined at each herd and parity group through the entire study period. The frequency of cows with BCS<3 or ∆BC>-0.5 at each cohort were calculated and used to define quartiles through whole study period. Quartiles were used, one at a time, as threshold to dichotomize the cohorts to predict the risk that a cohort has a frequency of anestrus over the median.The higher AUC was used as selection criterium to determine the herd level threshold at each HERD and PARITY level. The population attributable fraction (AFP) of anestrus rate to body condition indicators at each cohort was calculated, for every HERD and PARITY level.

Rotation Period Predictions for Low-mass Stars with Kinematic Information from Gaia DR3

<pre>DR3_kine_prot_pred.csv: 17.6 million predicted periods for all stars with Gaia DR3 RVs. This data should only be used to vet periods and not directly as period measurements. Works best for stars with GBP-GRP > 1.5 and rotation period > 20 days. Do not work for fast-rotating low-mass stars. Column descriptions see below </pre> source_id: Gaia DR3 source_id ra: Gaia DR3 ra measurements dec: Gaia DR3 dec measurements parallax: Gaia DR3 parallax measurements bp_rp: Gaia DR3 GBP-GRP measurements phot_g_mean_mag: Gaia DR3 G mag abs_G: absolute Gaia G magnitude derived from Gaia DR3 G mag and parallax ruwe: Gaia DR3 ruwe measurements  Prot_pred: Predicted periods using Gaia DR3 parameters and kinematics   vet_ztf_lowmass.csv: 65k vetted ZTF period measurements for stars with GBP-GRP > 1.5. Column descriptions see below source_id: Gaia DR3 source_id ra: Gaia DR3 ra measurements dec: Gaia DR3 dec measurements parallax: Gaia DR3 parallax measurements bp_rp: Gaia DR3 GBP-GRP measurements gmag: Gaia DR3 G mag abs_G: absolute Gaia G magnitude derived from Gaia DR3 G mag and parallax rv: Gaia DR3 rv measurements Prot: measured and vetted periods from ZTF

opencc-zeroJun 2023View details →

Data from: Brain criticality predicts individual levels of inter-areal synchronization in human electrophysiological data

Neuronal oscillations and their synchronization between brain areas are fundamental for healthy brain function. Yet, synchronization levels exhibit large inter-individual variability that is associated with behavioral variability. We test whether individual synchronization levels are predicted by individual brain states along an extended regime of critical-like dynamics – the Griffiths phase (GP). We use computational modelling to assess how synchronization is dependent on brain criticality indexed by long-range temporal correlations (LRTCs). We analyze LRTCs and synchronization of oscillations from resting-state magnetoencephalography and stereo-electroencephalography data. Synchronization and LRTCs are both positively linearly and quadratically correlated among healthy subjects, while in epileptogenic areas they are negatively linearly correlated. These results show that variability in synchronization levels is explained by the individual position along the GP with healthy brain areas operating in its subcritical and epileptogenic areas in its supercritical side. We suggest that the GP is fundamental for brain function allowing individual variability while retaining functional advantages of criticality.

PIGNet2: A versatile deep learning-based protein-ligand interaction prediction model for accurate binding affinity scoring and virtual screening

Training and test datasets of the paper "Improving the versatility of deep learning-based protein-ligand interaction prediction for accurate binding affinity scoring and virtual screening".

opencc-by-4.0May 2023View details →

Online course materials in predictive microbiology applied to growth - the EFSA ListeriaPredict Project

Online course materials in predictive microbiology applied to growth A deliverable from the EFSA funded ListeriaPredict Project

Data and codes: Speech-recognition in landlide predictive modelling

This is the data and codes for the manuscript "Speech-recognition in landlide predictive modelling"

opencc-zeroJul 2023View details →

Data from: Serum metabolomics predicts treatment response in myasthenia gravis

High-dose prednisone is the primary initial therapy for myasthenia gravis (MG) but upwards of a third of patients do not respond to treatment. No biomarkers can predict clinical responsiveness to corticosteroid treatment. We conducted a discovery-based study to identify treatment responsive biomarkers in MG using sera obtained at study entry to the thymectomy clinical trial (MGTX), an NIH-sponsored randomized, controlled study of thymectomy plus prednisone versus prednisone alone. We applied ultra-performance liquid chromatography coupled with electro-spray quadrupole time of flight mass spectrometry to obtain comparative serum metabolomic and lipidomic profiles at study entry to correlate with treatment response at 6 months. Treatment response was assessed using validated outcome measures of minimal manifestation status (MMS), MG-Activities of Daily Living (MG-ADL), Quantitative MG (QMG) score, or a strictly defined composite measure of response. Increased levels of phospholipids were associated with treatment response as assessed by QMG, MMS, and the Responders classification, but all measures showed limited overlap in metabolomic profiles, in particular the MG-ADL. A panel including histidine, free fatty acid (13:0), γ-cholestenol and guanosine was highly predictive of the strictly defined treatment response measure. The AUC in Responders' prediction for these markers was 0.90. Pathway analysis suggests that xenobiotic metabolism could play a major role in treatment resistance. We have defined a metabolomic and lipidomic profile that can now undergo validation as treatment predictive markers for MG patients undergoing corticosteroid therapy. Distinct metabolomic profiles were appreciated for each outcome measure.

opencc-zeroJul 2023View details →

Geospatial based model for malaria risk prediction in Kilombero Valley, south-eastern Tanzania

<div> Background: Malaria continues to pose a major public health challenge in tropical regions. Despite significant efforts to control malaria in Tanzania, there are still residual transmission cases. Unfortunately, little is known about where these residual malaria transmission cases occur and how they spread. In Tanzania, for example, the transmission is heterogeneously distributed. In order to effectively control and prevent the spread of malaria, it is essential to understand the spatial distribution and transmission patterns of the disease. This study seeks to predict areas that are at high risk of malaria transmission so that intervention measures can be developed to accelerate malaria elimination efforts. </div> Methods: This study employs a geospatial-based model to predict and map out malaria risk area in Kilombero Valley. Environmental factors related to malaria transmission were considered and assigned valuable weights in the Analytic Hierarchy Process (AHP), an online system using a pairwise comparison technique. The malaria hazard map was generated by a weighted overlay of the altitude, slope, curvature, aspect, rainfall distribution, and distance to streams in Geographic Information Systems (GIS). Finally, the risk map was created by overlaying components of malaria risk including hazards, elements at risk, and vulnerability. Results: The study demonstrates that the majority of the study area falls under the moderate-risk level (61%), followed by the low-risk level (31%), while the high-malaria risk area covers a small area, which occupies only 8% of the total area. Conclusion: The findings of this study are crucial for developing spatially targeted interventions against malaria transmission in residual transmission settings. Predicted areas prone to malaria risk provide information that will inform decision-makers and policymakers for proper planning, monitoring, and deployment of interventions.

opencc-by-4.0Feb 2023View details →

Alphafold predicted structures of VPS13 proteins from model organisms

This upload contains AlphaFold-predicted structures of VPS13 proteins from a variety of organisms. Given the large size of these proteins, only partial sequences were predicted with AlphaFold(1) and the resulting structures were aligned in PyMOL(2). A summary of the structures uploaded here is presented as a collection of domain cartoons in the "VPS13 domain organization across eukaryotic evolution.pdf" file.  The structures were generated with AlphaFold v2.029 on the Yale High Performance Cluster. Each *.zip file contains the best ranked predictions (out of five) for each sequence (*.pdb files) and the PyMOL assembled full structure (*.pse file). In a few cases, where a good alignment was not possible due to long disordered regions in the C-terminal portions (mostly in proteins from D. discoideum and A. thaliana), the full structures were aligned manually in PyMOL based on the continuity of the lipid transfer groove. The structures in PyMOL can be colour-coded by the confidence value of AlphaFold predictions using the following prompt: set_color n0, [0.051, 0.341, 0.827] set_color n1, [0.416, 0.796, 0.945] set_color n2, [0.996, 0.851, 0.212] set_color n3, [0.992, 0.490, 0.302] color n0, b < 100; color n1, b < 90 color n2, b < 70;  color n3, b < 50 Considering that full length structures were assembled by aligning different protein fragments and in view of the presence of flexible loops with low prediction confidence scores, the relative positions of different folded domains are not necessarily correct.   References 1. J. Jumper, et al., Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021). 2. The PyMOL Molecular Graphics System, Version 2.0. Schrödinger LLC.

Petascale Homology Search for Structure Prediction - MSAs

Multiple sequence alignments (MSAs) of CASP15 targets used in "Petascale Homology Search for Structure Prediction" publication.  Each MSA is constructed from a combination of different sequence databases, using different search tools, 1) ColabFoldDB (CFDB) using ColabFold search module, 2) Sequence Read Archive (SRA) using MMseqs2, and 3) UniRef30+BFD using HHblits (HH). <ul> <li>Queries: Regular CASP15 TS targets, excluding TBM-easy</li> <li>MSAs (A3M)  <ul> <li>CFDB (cfdb.tar.gz)</li> <li>SRA + CFDB (sra_cfdb.tar.gz)</li> <li>HH + CFDB (hh_cfdb.tar.gz)</li> <li>HH + SRA + CFDB (hh_sra_cfdb.tar.gz)</li> </ul> </li> </ul>

Predicted Spatially Complete Zoning Map of North Carolina

Spatially-complete zoning map of North Carolina, USA. The results folder contains results of a machine learning (random forest) model predicting 3 core district zones (residential, non-residential, and mixed use) and 13 sub-district zones (open space, industrial, commercial, office, planned use, high-density residential, medium-high-density residential, medium-density residential, medium-low-density residential, low-density residential, agricultural residential, mixed use, and downtown). Results are provided as 30-m rasters (.tif) with each value corresponding to a zoning district. Table containing zone district ID (number) and zone district name (character string) is included in zone_classification.csv. Final (spatially complete statewide maps) can be found in the final_predicted folder. This folder includes Statewide core district results in NC_predicted_core.tif and statewide sub-district results in NC_predicted_sub.tif.  Zoning was generalized and reclassified into 3 core district zones and 13 sub-district zones (described above). Reclassified zoning data, collected from 39 counties in North Carolina is provided in the observed folder with core districts in core_district_observed_zones.tif and sub-districts in sub_district_observed_zones.tif. Also in this folder is zoning_implementation_NC.csv which includes links to the source data (zoning map and zoning ordinance) for all collected data. Two models were created to predict zones under different data availability scenarios (i.e., scenarios that assume different levels of data availability). Predictions labeled “within_county” utilized the within-county model which predicts zoning districts in areas where zoning data is partially available for that county. To approximate scenarios of incomplete data accessibility, 20% of the data was randomly removed from training and reserved for independent performance assessments. Predictions labeled “between-county” utilized the between-county model which predicts zoning districts in areas where zoning data is inaccessible. To approximate this scenario, multiple between-county model iterations were computed by randomly removing entire counties from the training dataset and computing performance metrics on the removed (test) counties. Predictions are provided for both core districts and sub-districts (described above). Results from these models can be found in the predicted folder. This folder contains four subfolders: core_district_within_county, sub_district_within_county, core_district_between_county, and sub_district_between_county. Within each of these folders are predicted maps 30-m raster (.tif), performance reports including precision, recall, and f1 score overall and per district (.csv), and accuracy maps (3-km grid shapefile [.shp, .shx, .prj, .dbf]) with values corresponding to the proportion of misclassified pixels within a grid cell. Multiple randomized testing county samples were conducted for the between-county models. Each random sample is labeled r*_ where * is replaced with a number between 1 and 15.

Prediction of potential suitable areas for Phoebe zhennan in future different climate scenarios

This dataset includes sample collection data of existing Phoebe zhennan in China, as well as historical climate data and future climate data (with a resolution of 2.5 minutes and using the BCC-CSM2-MR GCM model) collected by Worldclim, along with geographical elevation data. These data are used to predict the potential distribution range of Phoebe zhennan in the future.

A multi-model ensemble of baseline and process-based models improves the predictive skill of near-term lake forecasts: data, forecasts, and scores

This data publication contains zipped parquet from the Falling Creek Reservoir multi-model ensemble (MME) forecasting work using the FLARE (Forecasting Lake And Reservoir Ecosystems) system and baseline models: drivers.zip contains NOAA driver forecast files, targets.zip contains in-situ water temperature observations, forecasts.zip contains forecast parquet files generated from the MME workflow (FLARE & baseline models), and scores.zip contains forecast skill metrics required for analysis.