Dataset Information

Fragment contribution models for predicting skin permeability using HuskinDB.

ABSTRACT: Mathematical models to predict skin permeation tend to be based on animal derived experimental data as well as knowing physicochemical properties of the compound under investigation, such as molecular volume, polarity and lipophilicity. This paper presents a strikingly contrasting model to predict permeability, formed entirely from simple chemical fragment (functional group) data and a recently released, freely accessible human (i.e. non-animal) skin permeation database, known as the 'Human Skin Database - HuskinDB'. Data from within the database allowed development of several fragment-based models, each including a calculable effect for all of the most commonly encountered functional groups present in compounds within the database. The developed models can be applied to predict human skin permeability (logK_p) for any compound containing one or more of the functional groups analysed from the dataset with no need to know any other physicochemical properties, solely the type and number of each functional group within the chemical structure itself. This approach simplifies mathematical prediction of permeability for compounds with similar properties to those used in this study.

SUBMITTER: Waters LJ

PROVIDER: S-EPMC10667307 | biostudies-literature | 2023 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Fragment contribution models for predicting skin permeability using HuskinDB.

Waters Laura J LJ Cooke David J DJ Quah Xin Ling XL

Scientific data 20231123 1

Mathematical models to predict skin permeation tend to be based on animal derived experimental data as well as knowing physicochemical properties of the compound under investigation, such as molecular volume, polarity and lipophilicity. This paper presents a strikingly contrasting model to predict permeability, formed entirely from simple chemical fragment (functional group) data and a recently released, freely accessible human (i.e. non-animal) skin permeation database, known as the 'Human Skin ...[more]

PMID: 37996523

Similar Datasets

Project description:The Biopharmaceutics Drug Disposition Classification System (BDDCS) can be utilized to predict drug disposition, including interactions with other drugs and transporter or metabolizing enzyme effects based on the extent of metabolism and solubility of a drug. However, defining the extent of metabolism relies upon clinical data. Drugs exhibiting high passive intestinal permeability rates are extensively metabolized. Therefore, we aimed to determine if in vitro measures of permeability rate or in silico permeability rate predictions could predict the extent of metabolism, to determine a reference compound representing the permeability rate above which compounds would be expected to be extensively metabolized, and to predict the major route of elimination of compounds in a two-tier approach utilizing permeability rate and a previously published model predicting the major route of elimination of parent drug. Twenty-two in vitro permeability rate measurement data sets in Caco-2 and MDCK cell lines and PAMPA were collected from the literature, while in silico permeability rate predictions were calculated using ADMET Predictor or VolSurf+. The potential for permeability rate to differentiate between extensively and poorly metabolized compounds was analyzed with receiver operating characteristic curves. Compounds that yielded the highest sensitivity-specificity average were selected as permeability rate reference standards. The major route of elimination of poorly permeable drugs was predicted by our previously published model, and the accuracies and predictive values were calculated. The areas under the receiver operating curves were >0.90 for in vitro measures of permeability rate and >0.80 for the VolSurf+ model of permeability rate, indicating they were able to predict the extent of metabolism of compounds. Labetalol and zidovudine predicted greater than 80% of extensively metabolized drugs correctly and greater than 80% of poorly metabolized drugs correctly in Caco-2 and MDCK, respectively, while theophylline predicted greater than 80% of extensively and poorly metabolized drugs correctly in PAMPA. A two-tier approach predicting elimination route predicts 72 ± 9%, 49 ± 10%, and 66 ± 7% of extensively metabolized, biliarily eliminated, and renally eliminated parent drugs correctly when the permeability rate is predicted in silico and 74 ± 7%, 85 ± 2%, and 73 ± 8% of extensively metabolized, biliarily eliminated, and renally eliminated parent drugs correctly when the permeability rate is determined in vitro.

Project description:In recent years, hydroponic greenhouse cultivation has gained increasing popularity: the combination of hydroponics' highly efficient use of resources with a controlled environment and an extended growing season provided by greenhouses allows for optimized, year-round plant growth. In this direction, precise and effective irrigation management is critical for achieving optimal crop yield while ensuring an economical use of water resources. This study explores techniques for explaining and predicting daily water consumption by utilizing only easily readily available meteorological data and the progressively growing records of the water consumption dataset. In situations where the dataset is limited in size, the conventional purely data-based approaches that rely on statistically benchmarking time series models tend to be too uncertain. Therefore, the objective of this study is to explore the potential contribution of crop models' main concepts in constructing more robust models, even when plant measurements are not available. Two strategies were developed for this purpose. The first strategy utilized the Greenlab model, employing reference parameter values from previously published papers and re-estimating, for identifiability reasons, only a limited number of parameters. The second strategy adopted key principles from crop growth models to propose a novel modeling approach, which involved deriving a Stochastic Segmentation of input Energy (SSiE) potentially absorbed by the elementary photosynthetically active parts of the plant. Several model versions were proposed and adjusted using the maximum likelihood method. We present a proof-of-concept of our methodology applied to the ekstasis Tomato, with one recorded time series of daily water uptake. This method provides an estimate of the plant's dynamic pattern of light interception, which can then be applied for the prediction of water consumption. The results indicate that the SSiE models could become valuable tools for extracting crop information efficiently from routine greenhouse measurements with further development and testing. This, in turn, could aid in achieving more precise irrigation management.

Project description:IntroductionAlthough clinical, functional, and biomarker data predict asthma exacerbations, newer approaches providing high accuracy of prognosis are needed for real-world decision-making in asthma. Machine learning (ML) leverages mathematical and statistical methods to detect patterns for future disease events across large datasets from electronic health records (EHR). This study conducted training and fine-tuning of ML algorithms for the real-world prediction of asthma exacerbations in patients with physician-diagnosed asthma.MethodsAdults with ≥ 2 ICD9/10 asthma codes within 1 year and at least 30 days apart were identified from the Optum Panther EHR database between 2016 and 2023. An emergency department (ED), urgent care, or inpatient visit for asthma, while on systemic administration of corticosteroids, was considered an exacerbation. To predict factors associated with exacerbations in a 6-month study period, clinical information from patients was retrieved in the preceding 6-month baseline period. Clinical information included demographics, lab results, diagnoses, medications, immunizations, and allergies. Three models built using Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and Transformers algorithms were trained and tested on independent datasets. Predictions were explained using the SHAP (SHapley Additive exPlanations) library.ResultsOf 1,331,934 patients with asthma, 16,279 (1.2%) experienced ≥ 1 exacerbation. XGBoost was the best predictive algorithm (area under the curve [AUC] = 0.964). Factors associated with exacerbations included a prior history of exacerbation, prednisone usage, high-dose albuterol usage, and elevated troponin I. Reduced probability of exacerbations was associated with receiving inhaled albuterol, vitamins, aspirin, statins, furosemide, and influenza vaccination.ConclusionThis ML-based study on asthma in the real world confirmed previously known features associated with increased exacerbation risk for asthma, while uncovering not entirely understood features associated with reduced risk of asthma exacerbations. These findings are hypothesis-generating and should contribute to ongoing discussion of the strengths and limitations of ML and other supervised learning models in patient risk stratification.

Dataset Information

Fragment contribution models for predicting skin permeability using HuskinDB.

Publications

Fragment contribution models for predicting skin permeability using HuskinDB.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets