Predicting hepatotoxicity using ToxCast in vitro bioactivity and chemical structure.

Ionic liquids (ILs) possess a unique physicochemical profile providing a wide range of applications. However, their “greenness”, specifically their claimed relative non toxicity has been frequently questioned, hindering their REACH registration processes and so, their final application. In this work we introduce a reliable, predictive, simple and chemically interpretable classification and regression tree (CART) classifier enabling the prioritization of ILs with a favourable cytotoxicity profile. By inspecting the structure of the CART several moieties that can be regarded as “cytotoxicophores” were identified and used to establish a set of SAR trends specifically aimed to prioritise low cytotoxicity ILs. We also demonstrated the suitability of the joint use of the CART classifier and a group fusion similarity search as a virtual screening strategy for the automatic prioritisation of safe ILs disperse in a data set of ILs of moderate to very high cytotoxicity. Additionally, we decided to complement the quantitative results already obtained by applying the network-like similarity graphs (NSG) approach to the mining of relevant structure-cytotoxicity relationships (SCR) trends. Finally, the SCR information concurrently gathered by both, quantitative (CART classifier) and qualitative (NSG) approaches was used to design a focused combinatorial library enriched with potentially safe ILs.

Background: Near-infrared spectroscopy (NIRS) has recognized potential but limited application for non-invasive diagnostic evaluation. Data analysis methodology that reproducibly distinguishes between the presence or absence of physiologic abnormality could broaden clinical application of this optical technique.Methods: Sample data sets from simultaneous NIRS bladder monitoring and invasive urodynamic pressure-flow studies (UDS) are used to illustrate how a diagnostic algorithm is constructed using classification and regression tree (CART) analysis. Misclassification errors of CART and linear discriminant analysis (LDA) are computed and examples of other urological NIRS data likely amenable to CART analysis presented.Results: CART generated a clinically relevant classification algorithm (error 4%) using 46 data sets of changes in chromophore concentration composed of the whole time series without specifying features. LDA did not (error 16%). Using CART NIRS data provided comparable discriminant ability to the UDS diagnostic nomogram for the presence or absence of obstructive pathology (88% specificity, 84% precision). Pilot data examples from children with and without voiding dysfunction and women with mild or severe pelvic floor muscle dysfunction also show potentially diagnostic differences in chromophore concentration.Conclusions: CART analysis can likely be applied in other NIRS monitoring applications intended to classify patients into those with and without pathology.

Background: Medication-related osteonecrosis of the jaw (MRONJ) is a rare but serious adverse event. Herein, we conducted a quantitative structure-activity relationship analysis using the U.S. Food and Drug Administration Adverse Drug Reaction Database System (FAERS) and machine learning to construct a drug prediction model for MRONJ induction based solely on chemical structure information. Methods: A total of 4815 drugs from FAERS were evaluated, including 70 and 139 MRONJ-positive and MRONJ-negative drugs, respectively, identified based on reporting odds ratios, Fisher's exact tests, and ≥100 total adverse event reports. Then, we calculated 326 chemical structure descriptors for each drug and compared three supervised learning algorithms (random forest, gradient boosting, and artificial neural networks). We also compared the number of chemical structure descriptors (5, 6, 7, 8, 9, 10, 20, and 30 descriptors). Results: We indicated that the MRONJ prediction model using an artificial neural network algorithm and eight descriptors achieved the highest validation receiver operating characteristic curve value of 0.778. Notably, the total polar surface area (ASA_P) was among the top-ranking descriptors, and MRONJ-positive drugs such as bisphosphonates and anticancer drugs showed high values. Our final model demonstrated a balanced accuracy of 0.693 and a specificity of 0.852. Conclusions: In this study, our MRONJ-inducing drug prediction model identified drugs with polar surface area properties as potential causes of MRONJ. This study demonstrates a promising approach for predicting MRONJ risk, which could enhance drug safety assessment and streamline drug screening in clinical and preclinical settings.

The growing human population accelerates alterations in land use and land cover (LULC) over time, putting tremendous strain on natural resources. Monitoring and assessing LULC change over large areas is critical in a variety of fields, including natural resource management and climate change research. LULC change has emerged as a critical concern for policymakers and environmentalists. As the need for the reliable estimation of LULC maps from remote sensing data grows, it is critical to comprehend how different machine learning classifiers perform. The primary goal of the present study was to classify LULC on the Google Earth Engine platform using three different machine learning algorithms—namely, support vector machine (SVM), random forest (RF), and classification and regression trees (CART)—and to compare their performance using accuracy assessments. The LULC of the study area was classified via supervised classification. For improved classification accuracy, NDVI (normalized difference vegetation index) and NDWI (normalized difference water index) indices were also derived and included. For the years 2016, 2018, and 2020, multitemporal Sentinel-2 and Landsat-8 data with spatial resolutions of 10 m and 30 m were used for the LULC classification. ‘Water bodies’, ‘forest’, ‘barren land’, ‘vegetation’, and ‘built-up’ were the major land use classes. The average overall accuracy of SVM, RF, and CART classifiers for Landsat-8 images was 90.88%, 94.85%, and 82.88%, respectively, and 93.8%, 95.8%, and 86.4% for Sentinel-2 images. These results indicate that RF classifiers outperform both SVM and CART classifiers in terms of accuracy.

Abstract. Vulnerability plays an important role in risk assessment. For flood vulnerability assessment, the map and characteristics of elements-at-risk at different scales are strongly required depending on the risk and vulnerability assessment requirements. This study proposes a methodology to classify urban structure type by combining object-based image classification and different high resolution remote sensing data. In this study, a high resolution satellite image and LiDAR have been acquired over Kota Bharu, Kelantan which consists of highly heterogeneous urban structure type (UST) classes. The first stage is data pre-processing that includes orthorectification and pansharpening of Geoeye satellite image, image resampling for normalised Digital Surface Model (nDSM) and followed by image segmentation for creating meaningful objects. The second stage comprises of derivation of image features, generation of training and testing datasets, and classification of UST. The classification was based on three types of machine learning classifiers, i.e. Random Forest (RF), Support Vector Machine (SVM) and Classification and Regression Tree (CART). The results obtained from the classification processes were compared using individual omission and commission error, overcall accuracy and Kappa coefficient. The results show that Random Forest classifier with all image features achieved the highest overall accuracy (93.5%) and Kappa coefficient (0.94). This is followed by CART classifier with overall accuracy of 93.7% and Kappa coefficient of 0.92. Finally, SVM classifier produced the lowest overall accuracy and Kappa coefficient with 88.6% and 0.86, respectively. The UST classification result can be further used to assist detailed building characterisation for large scale flood vulnerability assessment.

A comparison of different rule-based statistical models for modeling geogenic groundwater contamination

Quantitative Structure–Activity Relationship (QSAR) is an important computational technique used in modern drug discovery to predict the biological activity of chemical compounds based on their molecular structures. QSAR models establish mathematical relationships between chemical structure descriptors and biological activities. This method significantly reduces the cost, time, and resources required for experimental screening of large numbers of compounds. QSAR plays a major role in medicinal chemistry, toxicology, and environmental chemistry by helping researchers understand how structural features influence biological activity. The development of QSAR involves dataset preparation, calculation of molecular descriptors, model development, validation, and prediction. Modern QSAR approaches also integrate machine learning and artificial intelligence to improve prediction accuracy. This review article provides a clear and systematic overview of QSAR concepts, types of QSAR models, molecular descriptors, model development processes, validation techniques, applications, advantages, and limitations. The article is intended to provide fundamental knowledge of QSAR for learners and moderate-level researchers in the fields of medicinal chemistry and computational drug design.

The consistent rise of plastic pollution has stimulated interest in the development of biodegradable plastics. However, the study of polymer biodegradation has historically been limited to a small number of polymers due to costly and slow standard methods for measuring degradation, slowing new material innovation. High-throughput polymer synthesis and a high-throughput polymer biodegradation method are developed and applied to generate a biodegradation dataset for 642 chemically distinct polyesters and polycarbonates. The biodegradation assay was based on the clear-zone technique, using automation to optically observe the degradation of suspended polymer particles under the action of a single Pseudomonas lemoignei bacterial colony. Biodegradability was found to depend strongly on aliphatic repeat unit length, with chains less than 15 carbons and short side chains improving biodegradability. Aromatic backbone groups were generally detrimental to biodegradability; however, ortho- and para-substituted benzene rings in the backbone were more likely to be degradable than metasubstituted rings. Additionally, backbone ether groups improved biodegradability. While other heteroatoms did not show a clear improvement in biodegradability, they did demonstrate increases in biodegradation rates. Machine learning (ML) models were leveraged to predict biodegradability on this large dataset with accuracies over 82% using only chemical structure descriptors.

Recent advances in machine learning techniques have significantly influenced the development of early diabetes detection systems. In this paper, different machine learning (ML) algorithms are employed for diabetes detection based on hand movement. The spatial and temporal information for horizontal and vertical hand movement deviations for both diabetic and non-diabetic subjects are obtained using a data acquisition system. A dataset is constructed based on the obtained hand movement information along with other related attributes such as age, gender, weight and vision health. Binary classification of diabetic and non-diabetic subjects is achieved using several machine learning algorithms. These algorithms include Naïve Bayes (NB), Logistic Regression (LR), Linear Discriminate Analysis (LDA), Classification and Regression Tree (CART), k-Nearest Neighbors (KNN), Support Vector Machine (SVM) and Multi-layer Perceptron (MLP). The considered ML algorithms are evaluated and ranked according to a set of key performance metrics that include accuracy, precision, recall and F-1 score. Algorithm ranking results demonstrate that LR, CART, KNN and MLP algorithms have achieved higher ranking positions as compared to other algorithms. The best ranked ML algorithms are then validated and thoroughly investigated to prove their ability to discriminate between diabetic and non-diabetic subjects. Validation results show that the best-ranked ML algorithms have achieved high classification accuracy, precision, recall and F-1 scores. Therefore, the considered machine learning algorithms can serve as efficient tools for diabetes detection based on the constructed dataset.

In this paper, a complex morphological comparison of four Common Nightingale groups (Luscinia megarhynchos) is demonstrated. In total, 121 territorial nightingales were mist-netted and measured individually on four study areas called ‘Bódva’, ‘Felső-Tisza’, ‘Szatmár-Bereg’ and ‘Bátorliget’ in the North-Eastern part of Hungary in 2006–2013. To distinguish groups by morphology, Classification and Regression Trees (CART), Random Forest (RF) and Linear Discriminant Analysis (LDA) methods were used. Comparison of the four studied Common Nightingale groups shows substantial morphological differences in the length of the second, the third and the fourth primaries (P2, P3, P4), in bill length (BL) and bill width (BW), while other characteristics showed greater similarities. Based on the results of all the applied classification methods, birds originated from Szatmár-Bereg were clearly distinguishable from the others. The differences in morphology can be explained by interspecific competition or phenotypic plasticity resulting from the change of ecological, environmental parameters. Our case study highlights the advantageous differences of the classification methods to distinguish groups with similar morphology and to choose important variables for classification. In conclusion, broad application of the classification methods RF and CART is highly recommended in comparative ecological studies.

5 Finnish Game and Fisheries Research Institute, Evo State Fisheries and Aquaculture Research Station, Evo, Finland, SF-16970 Abstract. To evaluate the roles of extinction and isolation in predicting richness and composition of fish assemblages in small forest lakes of Finland and Wisconsin, we analyzed data from 114 Finnish and 55 Wisconsin lakes 0.2-86.9 ha in area. Six isolation variables characterized properties of stream corridors, land barriers, and source pools of invading species; four extinction variables were related to habitat severity, lake area, and productivity. Two types of multivariate analyses were used: the nonparametric classification and re- gression trees (CART) and the parametric linear discriminant analysis (LDA). Both types of analyses showed that extinction variables were collectively more important than isolation variables in predicting richness and composition both in Finland and Wis- consin. We interpret that the greater importance of extinction vs. isolation results, not because isolation is unimportant, but because the probability of an arrival of a new species is much less than that of an extinction. Thus, the time after an extinction event before a subsequent invasion is long relative to the time after an invasion event before a subsequent extinction; consequently, fish assemblages sampled at a given point in time more likely represent the stamp of the extinctions than of the invasions. This conclusion was robust to the differences in the geomorphic settings and fish faunas of Finland and Wisconsin. However, the importance of individual isolation and extinction variables in determining richness and composition differed between the two regions, apparently more from differ- ences in geomorphic settings than from differences in fish faunas. Influences of horizontal rather than vertical barriers over land and water were more apparent in Wisconsin, with its lower relief and higher incidence of lakes without stream connections; influences of the area of the nearest lake (representing the size of the available species pool) and stream gradient were more important in Finland, with its higher relief and higher incidence of lakes with stream connections. The importance of individual extinction variables also dif- fered between the two regions, again reflecting differences in the geomorphic settings of the two lake districts and the strong influence that lake position in the landscape has in determining limnological features of the lake.

Introducing two Random Forest based methods for cloud detection in remote sensing images

Evaluation of Classification Techniques for Arrhythmia Screening of Astronauts

BackgroundCommunity-based health insurance (CBHI) is a vital tool for achieving universal health coverage (UHC), a key global health priority outlined in the sustainable development goals (SDGs). Sub-Saharan Africa continues to face challenges in achieving UHC and protecting individuals from the financial burden of disease. As a result, CBHI has become popular in low- and middle-income countries, including Ethiopia. Therefore, this study aimed to identify the ML algorithm with the best predictive accuracy for CBHI enrollment and to determine the most influential predictors among the dataset.MethodsThe 2019 Ethiopian Mini Demographic and Health Survey (EMDHS) data were used. The CBHI were predicted using seven machine learning models: linear discriminant analysis (LDA), support vector machine with radial basis function (SVM), k-nearest neighbors (KNN), classification and regression tree (CART), and random forest (RF). Receiver operating characteristic curves and other metrics were used to evaluate each model’s accuracy.ResultsThe RF algorithm was determined to be the best machine learning model based on different performance assessments. The result indicates that age, wealth index, household members, and land usage all significantly affect CBHI in Ethiopia.ConclusionThis study found that RF machine learning models could improve the ability to classify CBHI in Ethiopia with high accuracy. Age, wealth index, household members, and land utilization are some of the most significant variables associated with CBHI that were determined by feature importance. The results of the study can help health professionals and policymakers create focused strategies to improve CBHI enrollment in Ethiopia.

Derin Öğrenme modelleri kompleks deneyimlerden bilgi çıkarımına imkân sağlayan spesifik Makine Öğrenmesi yöntemleridir. Kapalı bir mekândaki bazı veri değerlerindeki değişiminin öğrenilmesi ile odada herhangi bir kişinin bulunup bulunmamasının tespit edilmesi bu deneyimlerden biridir. Bu çalışmanın amacı zaman serileri olarak oluşturulmuş ve zaman içinde ışık, sıcaklık, nem ve CO2 değerlerindeki değişimler ile kapalı bir mekânda doluluk tespiti probleminin Bir Boyutlu Evrişimli Sinir Ağı (1D-CNN) ile gerçekleştirilmesidir. Bir adet eğitim ve iki adet test veri seti kullanılarak model eğitilmiş ve daha önce tecrübe edilmeyen test veri setleri ile modelin başarısı gözlenmiştir. Keras uygulama programlama arayüzünde 1D-CNN modeli ile gerçekleştirilen testlerde doluluk tespiti sınıflandırmasının RF (Random Forest), GBM (Gradient Boosting Machines), CART (Classification and Regression Trees), LDA (Linear Discriminant Analysis) yöntemlerinden daha başarılı sonuçlar verdiği gözlenmiştir.