Structure-Activity Relationship (SAR) Model for Predicting Teratogenic Risk of Antiseizure Medications in Pregnancy by Using Support Vector Machine

Abstract

Teratogenicity is one of the main concerns in clinical medications of pregnant women. Prescription of antiseizure medications (ASMs) in women with epilepsy during pregnancy may cause teratogenic effects on the fetus. Although large scale epilepsy pregnancy registries played an important role in evaluating the teratogenic risk of ASMs, for most ASMs, especially the newly approved ones, the potential teratogenic risk cannot be effectively assessed due to the lack of evidence. In this study, the analyses are performed on any medication, with a focus on ASMs. We curated a list containing the drugs with potential teratogenicity based on the US Food and Drug Administration (FDA)-approved drug labeling, and established a support vector machine (SVM) model for detecting drugs with high teratogenic risk. The model was validated by using the post-marketing surveillance data from US FDA Spontaneous Adverse Events Reporting System (FAERS) and applied to the prediction of potential teratogenic risk of ASMs. Our results showed that our proposed model outperformed the state-of-art approaches, including logistic regression (LR), random forest (RF) and extreme gradient boosting (XGBoost), when detecting the high teratogenic risk of drugs (MCC and recall rate were 0.312 and 0.851, respectively). Among 196 drugs with teratogenic potential reported by FAERS, 136 (69.4%) drugs were correctly predicted. For the eight commonly used ASMs, 4 of them were predicted as high teratogenic risk drugs, including topiramate, phenobarbital, valproate and phenytoin (predicted probabilities of teratogenic risk were 0.69, 0.60 0.59, and 0.56, respectively), which were consistent with the statement in FDA-approved drug labeling and the high reported prevalence of teratogenicity in epilepsy pregnancy registries. In addition, the structural alerts in ASMs that related to the genotoxic carcinogenicity and mutagenicity, idiosyncratic adverse reaction, potential electrophilic agents and endocrine disruption were identified and discussed. Our findings can be a good complementary for the teratogenic risk assessment in drug development and facilitate the determination of pharmacological therapies during pregnancy.

Advancing Adverse Drug Reaction Prediction with Deep Chemical Language Model for Drug Safety Evaluation

“…In previous studies, Wulin Long et al indicated that SVM achieved the highest MCC value and recall rate (mean MCC = 0.591, mean recall rate = 0.870), predicting the high risk of inducing QT interval prolongation of marketed drugs [13]. Liyuan Kang et al indicated that SVM performed better in detecting the marketed drugs with high teratogenic risk (mean MCC = 0.312) [12]. Yifan Zhou et al proposed that the RF model performed the best (mean MCC = 0.46) in predicting the DIR severity of the marketed drugs [16].…”

“…We utilized three meticulously curated datasets as follows: the drug-induced QT prolongation atlas (DIQTA, https://www.adratlas.com/DIQTA/download accessed on 6 February 2024) [22], the drug-induced teratogenicity dataset (DITD, https://www.frontiersin.org/articles/10.3389/fphar.2022.747935/full#supplementary-material accessed on 6 February 2024) [12], and the drug-induced rhabdomyolysis atlas (DIRA, https://www.adratlas.com/DIRA/download accessed on 6 February 2024) [23] for model construction and validation. All drugs in these datasets were classified into different concern levels based on the severity outlined in their FDA-approved drug labeling.…”

“…These methodologies offer pivotal approaches to correlate the chemical structure of drugs with their propensity for ADRs, representing indispensable tools for predicting drug-induced adverse reactions [9,10]. Previous research has demonstrated the utility of SAR models employing conventional machine-learning methods, including Support Vector Machine (SVM) [11][12][13][14], Random Forest (RF) [15][16][17], and Decision Forest (DF) [18,19]. Notably, these models have exhibited promising performance in predicting liver toxicity [15,[18][19][20], cardiotoxicity [13,14,17,21], and other adverse reactions.…”

“…Notably, these models have exhibited promising performance in predicting liver toxicity [15,[18][19][20], cardiotoxicity [13,14,17,21], and other adverse reactions. In our previous investigation, we constructed datasets encompassing three distinct ADRs based on FDA-approved drug labeling information: the drug-induced QT prolongation atlas (DIQTA) [22], drug-induced teratogenicity dataset (DITD) [12], and drug-induced rhabdomyolysis atlas (DIRA) [23]. Leveraging conventional machine-learning algorithms in conjunction with molecular descriptors, we predicted the risk associated with these three types of ADRs [12,13,16].…”

Predicting Fetal Alcohol Spectrum Disorders Using Machine Learning Techniques: Multisite Retrospective Cohort Study

“…As of now, few studies have attempted to use ML strategies to detect prenatal exposure to alcohol, despite its increased use in retrospective studies of other teratogens like tobacco [ 11 , 12 ], environmental contaminants [ 13 ], and certain medications [ 14 ]. For example, in an American cohort study of 531 children between 3 and 5 years old, ML algorithms achieved an accuracy of 81% in detecting prenatal exposure to smoking, by incorporating DNA methylation data and maternal self-reports [ 12 ].…”