ABSTRACT
Anatomical Therapeutic Chemical (ATC) classes prediction is one of the prominent activities in the costly and tedious pipeline of drug discovery where machine learning plays an important role by minimizing the cost and time of prediction. Most of the existing research have been done to predict ATC classes from the chemical-chemical association, side-effects, target proteins, gene expressions, chemical structures, drug targets, and textual information of drugs. However, the capability of 17 molecules’ properties have not yet been explored to predict drug ATC classes. The current work proposes a methodology for predicting the drug ATC classes using the 17 molecules’ properties. ATC classes prediction is a multi-label classification task and therefore, a binary relevance strategy has been employed to solve this issue with four basic machine learning classifiers, namely K-Nearest Neighbour (KNN), Extra Tree Classifier (ETC), Random Forest (RF), and Decision Tree (DT). The common problem of multi-label datasets is class imbalance which is addressed using the MLSMOTE (Multi-Label Synthetic Minority Over-Sampling Technique). The proposed methodology exhibits promising results, and it achieved the accuracy ranging from 96.90% to 98.06%, which indicates that 17 molecules’ properties are good enough in efficient prediction of ATC classes.
- A breakthrough milestone in AI-powered drug discovery reached linking biology and chemistry with AI. url: https://insilico.com/blog/pcc (visited on 07/12/2021).Google Scholar
- Andrew P Bradley. “The use of the area under the ROC curve in the evaluation of machine learning algorithms”. In: Pattern recognition 30.7 (1997), pp. 1145–1159.Google ScholarDigital Library
- Francisco Charte “MLSMOTE: Approaching imbalanced multilabel learning through synthetic instance generation”. In: Knowledge-Based Systems 89 (2015), pp. 385–397.Google ScholarDigital Library
- Fan-Shu Chen and Zhen-Ran Jiang. “Prediction of drug's Anatomical Therapeutic Chemical (ATC) code by integrating drug–domain network”. In: Journal of biomedical informatics 58 (2015), pp. 80–88.Google ScholarDigital Library
- Lei Chen, Tao Liu, and Xian Zhao. “Inferring anatomical therapeutic chemical (ATC) class of drugs using shortest path and random walk with restart algorithms”. In: Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease 1864.6 (2018), pp. 2228–2240.Google ScholarCross Ref
- Lei Chen “Predicting anatomical therapeutic chemical (ATC) classification of drugs by integrating chemical-chemical interactions and similarities”. In: PloS one 7.4 (2012), e35254.Google ScholarCross Ref
- etc.ensemble.ExtraTreesClassifier. url: https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.ExtraTreesClassif ier.html#sklearn.ensemble.ExtraTreesClassif ier (visited on08/27/2021).Google Scholar
- Harsha Gurulingappa “Concept-based semi-automatic classification of drugs”. In: Journal of chemical information and modeling 49.8 (2009), pp. 1986–1992.Google ScholarCross Ref
- Jeonghee Jo, Hyun-Soo Choi, and Sungroh Yoon. “Prediction of Drug Classes with a Deep Neu- ral Network using Drug Targets and Chemical Structure Data”. In: 2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM). IEEE. 2019, pp. 664–667.Google ScholarCross Ref
- Michael Kuhn “A side effect resource to capture phenotypic effects of drugs”. In: Molecular systems biology 6.1 (2010), p. 343.Google ScholarCross Ref
- Haiyan Liang “Recognizing novel chemicals/drugs for anatomical therapeutic chemical classes with a heat diffusion algorithm”. In: Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease 1866.11 (2020), p. 165910.Google ScholarCross Ref
- Zhongyang Liu “Similarity-based prediction for anatomical therapeutic chemical clas- sification of drugs by integrating multiple data sources”. In: Bioinformatics 31.11 (2015), pp. 1788–1795.Google ScholarCross Ref
- Gjorgji Madjarov “An extensive experimental comparison of methods for multi-label learning”. In: Pattern recognition 45.9 (2012), pp. 3084–3104.Google ScholarDigital Library
- Ken McGarry and Ennock Assamoha. “Data integration with self-organising neural network reveals chemical structure and therapeutic effects of drug ATC codes”. In: UK Workshop on Computational Intel ligence. Springer. 2017, pp. 63–74.Google Scholar
- Loris Nanni, Sheryl Brahnam, and Alessandra Lumini. “Ensemble of Deep Learning Approaches for ATC Classification”. In: Smart Intel ligent Computing and Applications. Springer, 2020, pp. 117–125.Google ScholarCross Ref
- Janette Nickel “SuperPred: update on drug classification and target prediction”. In: Nucleic acids research 42.W1 (2014), W26–W31.Google ScholarCross Ref
- Thomas Olson and Rahul Singh. “Predicting anatomic therapeutic chemical classification codes using tiered learning”. In: BMC bioinformatics 18.8 (2017), pp. 1–13.Google Scholar
- Yayuan Peng “Drug repositioning by prediction of drug's anatomical therapeutic chemical code via network-based inference approaches”. In: Briefings in bioinformatics 22.2 (2021), pp. 2058–2072.Google ScholarCross Ref
- Purvi Prajapati, Amit Thakkar, and Amit Ganatra. “A survey and current research challenges in multi-label classification methods”. In: International Journal of Soft Computing and Engineering (IJSCE) 2.1 (2012), pp. 248–252.Google Scholar
- PubChem. url: https://pubchem.ncbi.nlm.nih.gov/ (visited on 09/29/2021).Google Scholar
- Jesse Read “Classifier chains for multi-label classification”. In: Machine learning 85.3 (2011), pp. 333–359.Google ScholarDigital Library
- sklearn.ensemble.RandomForestClassifier. url: https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassif ier.html#sklearn.ensemble.RandomForestClass if ier (visited on 08/27/2021).Google Scholar
- sklearn.neighbors.KNeighborsClassifier. url: https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassif ier.html#sklearn.neighbors.KNeighborsClassif ier (visited on 08/27/2021).Google Scholar
- sklearn.tree.DecisionTreeClassifier. url: https://scikit- learn.org/stable/modules/generated/sklearn.tree.DecisionTreeClassif ier.html#sklearn.tree.DecisionTreeClassif ier (visited on 08/27/2021).Google Scholar
- Pankaj Vaidya, Ankit Gupta, and Varun Jaiswal. “Machine learning based prediction of anatomical therapeutic chemical (ATC) class of drug like molecule”. In: 2018 International Conference on Recent Innovations in Electrical, Electronics & Communication Engineering (ICRIEECE). IEEE. 2018, pp. 1045–1048.Google ScholarCross Ref
- Jessica Vamathevan “Applications of machine learning in drug discovery and develop- ment”. In: Nature Reviews Drug Discovery 18.6 (2019), pp. 463–477.Google ScholarCross Ref
- Chi-Shiang Wang “Detecting potential adverse drug reactions using a deep neural network model”. In: Journal of medical Internet research 21.2 (2019), e11016.Google ScholarCross Ref
- Xiangeng Wang “ATC-NLSP: prediction of the classes of anatomical therapeutic chemicals using a network-based label space partition method”. In: Frontiers in pharmacology 10 (2019), p. 971.Google ScholarCross Ref
- Yong-Cui Wang “Network predicting drug's anatomical therapeutic chemical code”. In: Bioinformatics 29.10 (2013), pp. 1317–1324.Google ScholarDigital Library
- Min-Ling Zhang “Binary relevance for multi-label learning: an overview”. In: Frontiers of Computer Science 12.2 (2018), pp. 191–202.Google ScholarDigital Library
- Haochen Zhao “A novel approach based on deep residual learning to predict drug's anatomical therapeutic chemical code”. In: 2020 IEEE International Conference on Bioinfor- matics and Biomedicine (BIBM). IEEE. 2020, pp. 921–926.Google ScholarCross Ref
- Jian-Peng Zhou, Lei Chen, and Zi-Han Guo. “iATC-NRAKEL: an efficient multi-label classifier for recognizing anatomical therapeutic chemical classes of drugs”. In: Bioinformatics 36.5 (2020), pp. 1391–1396.Google ScholarCross Ref
Recommendations
Network predicting drug’s anatomical therapeutic chemical code
Motivation: Discovering drug’s Anatomical Therapeutic Chemical (ATC) classification rules at molecular level is of vital importance to understand a vast majority of drugs action. However, few studies attempt to annotate drug’s potential ATC-codes by ...
Discovery of novel influenza inhibitors targeting the interaction of dsRNA with the NS1 protein by structure-based virtual screening
Influenza A Non-structural protein 1 (NS1A) RNA-Binding Domain (RBD) bound to a double-stranded RNA (dsRNA), which can inhibit the activation of antiviral pathway. The chemical compound binding sites at this pocket have abilities to block NS1 protein to ...
Using SMOTE to Deal with Class-Imbalance Problem in Bioactivity Data to Predict mTOR Inhibitors
AbstractMachine learning algorithms give sub-optimal performance in the presence of class-imbalanced dataset. Mammalian target of rapamycin (mTOR) is one of the serine/threonine protein kinase, and plays an integral role in autophagy pathway. Autophagy is ...
Comments