QSAR Models for Human Carcinogenicity: An Assessment Based on Oral and Inhalation Slope Factors

Unlocking the potential of AI: Machine learning and deep learning models for predicting carcinogenicity of chemicals

The escalating apprehension surrounding the carcinogenic potential of chemicals emphasizes the imperative need for efficient methods of assessing carcinogenicity. Conventional experimental approaches such as in vitro and in vivo assays, albeit effective, suffer from being costly and time-consuming. In response to this challenge, new alternative methodologies, notably machine learning and deep learning techniques, have attracted attention for their potential in developing carcinogenicity prediction models. This article reviews the progress in predicting carcinogenicity using various machine learning and deep learning algorithms. A comparative analysis on these developed models reveals that support vector machine, random forest, and ensemble learning are commonly preferred for their robustness and effectiveness in predicting chemical carcinogenicity. Conversely, models based on deep learning algorithms, such as feedforward neural network, convolutional neural network, graph convolutional neural network, capsule neural network, and hybrid neural networks, exhibit promising capabilities but are limited by the size of available carcinogenicity datasets. This review provides a comprehensive analysis of current machine learning and deep learning models for carcinogenicity prediction, underscoring the importance of high-quality and large datasets. These observations are anticipated to catalyze future advancements in developing effective and generalizable machine learning and deep learning models for predicting chemical carcinogenicity.

New approach methodologies to facilitate and improve the hazard assessment of non-genotoxic carcinogens-a PARC project

Carcinogenic chemicals, or their metabolites, can be classified as genotoxic or non-genotoxic carcinogens (NGTxCs). Genotoxic compounds induce DNA damage, which can be detected by an established in vitro and in vivo battery of genotoxicity assays. For NGTxCs, DNA is not the primary target, and the possible modes of action (MoA) of NGTxCs are much more diverse than those of genotoxic compounds, and there is no specific in vitro assay for detecting NGTxCs. Therefore, the evaluation of the carcinogenic potential is still dependent on long-term studies in rodents. This 2-year bioassay, mainly applied for testing agrochemicals and pharmaceuticals, is time-consuming, costly and requires very high numbers of animals. More importantly, its relevance for human risk assessment is questionable due to the limited predictivity for human cancer risk, especially with regard to NGTxCs. Thus, there is an urgent need for a transition to new approach methodologies (NAMs), integrating human-relevant in vitro assays and in silico tools that better exploit the current knowledge of the multiple processes involved in carcinogenesis into a modern safety assessment toolbox. Here, we describe an integrative project that aims to use a variety of novel approaches to detect the carcinogenic potential of NGTxCs based on different mechanisms and pathways involved in carcinogenesis. The aim of this project is to contribute suitable assays for the safety assessment toolbox for an efficient and improved, internationally recognized hazard assessment of NGTxCs, and ultimately to contribute to reliable mechanism-based next-generation risk assessment for chemical carcinogens.

Data-Driven Quantitative Structure-Activity Relationship Modeling for Human Carcinogenicity by Chronic Oral Exposure

Traditional methodologies for assessing chemical toxicity are expensive and time-consuming. Computational modeling approaches have emerged as low-cost alternatives, especially those used to develop quantitative structure-activity relationship (QSAR) models. However, conventional QSAR models have limited training data, leading to low predictivity for new compounds. We developed a data-driven modeling approach for constructing carcinogenicity-related models and used these models to identify potential new human carcinogens. To this goal, we used a probe carcinogen dataset from the US Environmental Protection Agency's Integrated Risk Information System (IRIS) to identify relevant PubChem bioassays. Responses of 25 PubChem assays were significantly relevant to carcinogenicity. Eight assays inferred carcinogenicity predictivity and were selected for QSAR model training. Using 5 machine learning algorithms and 3 types of chemical fingerprints, 15 QSAR models were developed for each PubChem assay dataset. These models showed acceptable predictivity during 5-fold cross-validation (average CCR = 0.71). Using our QSAR models, we can correctly predict and rank 342 IRIS compounds' carcinogenic potentials (PPV = 0.72). The models predicted potential new carcinogens, which were validated by a literature search. This study portends an automated technique that can be applied to prioritize potential toxicants using validated QSAR models based on extensive training sets from public data resources.

Application of a developed triple-classification machine learning model for carcinogenic prediction of hazardous organic chemicals to the US, EU, and WHO based on Chinese database

Cancer, the second largest human disease, has become a major public health problem. The prediction of chemicals' carcinogenicity before their synthesis is crucial. In this paper, seven machine learning algorithms (i.e., Random Forest (RF), Logistic Regression (LR), Support Vector Machines (SVM), Complement Naive Bayes (CNB), K-Nearest Neighbor (KNN), XGBoost, and Multilayer Perceptron (MLP)) were used to construct the carcinogenicity triple classification prediction (TCP) model (i.e., 1A, 1B, Category 2). A total of 1444 descriptors of 118 hazardous organic chemicals were calculated by Discovery Studio 2020, Sybyl X-2.0 and PaDEL-Descriptor software. The constructed carcinogenicity TCP model was evaluated through five model evaluation indicators (i.e., Accuracy, Precision, Recall, F1 Score and AUC). The model evaluation results show that Accuracy, Precision, Recall, F1 Score and AUC evaluation indicators meet requirements (greater than 0.6). The accuracy of RF, LR, XGBoost, and MLP models for predicting carcinogenicity of Category 2 is 91.67%, 79.17%, 100%, and 100%, respectively. In addition, the constructed machine learning model in this study has potential for error correction. Taking XGBoost model as an example, the predicted carcinogenicity level of 1,2,3-Trichloropropane (96-18-4) is Category 2, but the actual carcinogenicity level is 1B. But the difference between Category 2 and 1B is only 0.004, indicating that the XGBoost is one optimum model of the seven constructed machine learning models. Besides, results showed that functional groups like chlorine and benzene ring might influence the prediction of carcinogenic classification. Therefore, considering functional group characteristics of chemicals before constructing the carcinogenicity prediction model of organic chemicals is recommended. The predicted carcinogenicity of the organic chemicals using the optimum machine leaning model (i.e., XGBoost) was also evaluated and verified by the toxicokinetics. The RF and XGBoost TCP models constructed in this paper can be used for carcinogenicity detection before synthesizing new organic substances. It also provides technical support for the subsequent management of organic chemicals.

Predicting Chemical Carcinogens Using a Hybrid Neural Network Deep Learning Method

Determining environmental chemical carcinogenicity is urgently needed as humans are increasingly exposed to these chemicals. In this study, we developed a hybrid neural network (HNN) method called HNN-Cancer to predict potential carcinogens of real-life chemicals. The HNN-Cancer included a new SMILES feature representation method by modifying our previous 3D array representation of 1D SMILES simulated by the convolutional neural network (CNN). We developed binary classification, multiclass classification, and regression models based on diverse non-congeneric chemicals. Along with the HNN-Cancer model, we developed models based on the random forest (RF), bootstrap aggregating (Bagging), and adaptive boosting (AdaBoost) methods for binary and multiclass classification. We developed regression models using HNN-Cancer, RF, support vector regressor (SVR), gradient boosting (GB), kernel ridge (KR), decision tree with AdaBoost (DT), KNeighbors (KN), and a consensus method. The performance of the models for all classifications was assessed using various statistical metrics. The accuracy of the HNN-Cancer, RF, and Bagging models were 74%, and their AUC was ~0.81 for binary classification models developed with 7994 chemicals. The sensitivity was 79.5% and the specificity was 67.3% for the HNN-Cancer, which outperforms the other methods. In the case of multiclass classification models with 1618 chemicals, we obtained the optimal accuracy of 70% with an AUC 0.7 for HNN-Cancer, RF, Bagging, and AdaBoost, respectively. In the case of regression models, the correlation coefficient (R) was around 0.62 for HNN-Cancer and RF higher than the SVM, GB, KR, DTBoost, and NN machine learning methods. Overall, the HNN-Cancer performed better for the majority of the known carcinogen experimental datasets. Further, the predictive performance of HNN-Cancer on diverse chemicals is comparable to the literature-reported models that included similar and less diverse molecules. Our HNN-Cancer could be used in identifying potentially carcinogenic chemicals for a wide variety of chemical classes.

QSAR Studies on Nitrobenzene Derivatives using Hyperpolarizability and Conductor like Screening model as Molecular Descriptors

Quantitative structure-activity relationship (QSAR) models were useful in understanding how chemical structure relates to the toxicology of chemicals. In the present study, we report quantum molecular descriptors using conductor like screening model (COs) area, the linear polarizability, first and second order hyperpolarizability for modelling the toxicology of the nitro substituent on the benzene ring. All the molecular descriptors were performed using semi-empirical PM6 approaches. The QSAR model was developed using stepwise multiple linear regression. We found that the stable QSAR modelling of toxicology benzene derivatives used second order hyper-polarizability and COs area, which satisfied the statistical measures. The second order hyperpolarizability shows the best QSAR model. We also discovered that the nitrobenzene derivative’s substitutional functional group has a significant effect on the quantum molecular descriptors, which reflect the QSAR model.

Carcinogenicity prediction using the index of ideality of correlation

Carcinogenicity testing is necessary to protect human health and comply with regulations, but testing it with the traditionally used two-year rodent studies is time-consuming and expensive. In certain cases, such as for impurities, alternative methods may be convenient. Thus there is an urgent need for alternative approaches for reliable and robust assessments of carcinogenicity. The Monte Carlo technique with CORAL software is a tool to tackle this task for unknown compounds using available experimental data for a representative set of compounds. The models can be constructed with the simplified molecular input line entry system without additional physicochemical descriptors. We describe here a model based on a data set of 1167 substances. Matthew's correlation coefficient values for calibration and validation sets are 0.747 and 0.577, respectively. Double bonds between carbon atoms and double bonds of oxygen atoms are the molecular features that indicate the carcinogenic potential of a compound.

QSAR modelling of inhalation toxicity of diverse volatile organic molecules using no observed adverse effect concentration (NOAEC) as the endpoint

Nowadays, air pollution due to urbanization and reduction of forestry is emerging as a serious threat to humans and the environment. According to the World Health Organization, respiratory diseases are the third most mortality factor in the world. Chemical research organizations and industries are producing a large number of new chemical compounds continuously. Although toxicity testing of those chemicals on animals is costly, resource and time consuming, these data cannot be properly extrapolated to humans and other animals, and also these raise ethical issues. In this background, we have developed Quantitative Structure-Activity Relationship (QSAR) models using the No Observed Adverse Effect Concentration (NOAEC) as the endpoint to assess inhalation toxicity of diverse organic chemicals, commonly used and exposed by us in our daily life. No Observed Adverse Effect Concentration (NOAEC) can be used for long term toxicity studies towards the human inhalation risk assessment, as recommended by Organization for Economic Co-operation and Development (OECD) in guidance document 39. A particular QSAR model may not be equally effective for prediction of all query compounds from a given set of compounds; therefore, we have developed multiple models, which are robust, sound and well predictive from the statistical point of view to forecast the NOAEC values for the new untested compounds. Subsequently the validated individual models were employed to generate consensus models, in order to improve the quality of predictions and to reduce prediction errors. We have investigated some crucial structural features from these models which may regulate inhalation toxicity for newly produced molecules. Thus, our developed models may help in toxicity assessment towards reducing the health hazards for new chemicals.

In Silico Methods for Carcinogenicity Assessment

Screening compounds for potential carcinogenicity is of major importance for prevention of environmentally induced cancers. A large sequence of predictive models, ranging from short-term biological assays (e.g., mutagenicity tests) to theoretical models, has been attempted in this field. Theoretical approaches such as (Q)SAR are highly desirable for identifying carcinogens, since they actively promote the replacement, reduction, and refinement of animal tests. This chapter reports and describes some of the most noted (Q)SAR models based on human expert knowledge and statistical approaches, aiming at predicting the carcinogenicity of chemicals. Additionally, the performance of the selected models has been evaluated, and the results are interpreted in details by applying these predictive models to some pharmaceutical molecules.

Preoperative prediction of lymph node metastasis in patients with papillary thyroid carcinoma by an artificial intelligence algorithm

BACKGROUND

It is necessary to identify patients at risk of developing lymph node metastasis prior to papillary thyroid carcinoma (PTC) surgery. This can be challenging due to limiting factors, and an artificial intelligence algorithm may be a viable option.

OBJECTIVE

In this study, we aimed to evaluate whether combining an artificial intelligence algorithm (support vector machine and probabilistic neural network) and clinico-pathologic data can preoperatively predict lymph node metastasis of papillary thyroid carcinoma (PTC).

METHODS

We retrospectively examined 251 PTCs with lymph node metastasis and 194 PTCs without lymph node metastasis. The artificial intelligence algorithm included the support vector machine (SVM) and the probabilistic neural network (PNN).

RESULTS

The ACR TI-RADS (Thyroid Imaging, Reporting and Data System), number of tumours, no well-defined margin, lymph node status and rim calcification on ultrasonography (US), age, sex, tumour size, and presence of Hashimoto's thyroiditis were significantly more frequent among PTCs with central lymph node metastasis than those without metastasis (P<0.05). The PNN classifier revealed an F1 score of 0.88 on the central lymph node metastasis test set. The SVM classifier revealed an F1 score of 0.93 on the lateral lymph node metastasis test set. Our study demonstrates that combining artificial intelligence algorithms and clinico-pathologic data can effectively predict the lymph node metastasis of papillary thyroid carcinoma prior to surgery.

Defining the Human-Biota Thresholds of Toxicological Concern for Organic Chemicals in Freshwater: The Proposed Strategy of the LIFE VERMEER Project Using VEGA Tools

Several tons of chemicals are released every year into the environment and it is essential to assess the risk of adverse effects on human health and ecosystems. Risk assessment is expensive and time-consuming and only partial information is available for many compounds. A consolidated approach to overcome this limitation is the Threshold of Toxicological Concern (TTC) for assessment of the potential health impact and, more recently, eco-TTCs for the ecological aspect. The aim is to allow a safe assessment of substances with poor toxicological characterization. Only limited attempts have been made to integrate the human and ecological risk assessment procedures in a "One Health" perspective. We are proposing a strategy to define the Human-Biota TTCs (HB-TTCs) as concentrations of organic chemicals in freshwater preserving both humans and ecological receptors at the same time. Two sets of thresholds were derived: general HB-TTCs as preliminary screening levels for compounds with no eco- and toxicological information, and compound-specific HB-TTCs for chemicals with known hazard assessment, in terms of Predicted No effect Concentration (PNEC) values for freshwater ecosystems and acceptable doses for human health. The proposed strategy is based on freely available public data and tools to characterize and group chemicals according to their toxicological profiles. Five generic HB-TTCs were defined, based on the ecotoxicological profiles reflected by the Verhaar classes, and compound-specific thresholds for more than 400 organic chemicals with complete eco- and toxicological profiles. To complete the strategy, the use of in silico models is proposed to predict the required toxicological properties and suitable models already available on the VEGAHUB platform are listed.