{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deepcci-end-to-end-deep-learning-for-chemical","title":"DeepCCI: End-to-end Deep Learning for Chemical-Chemical Interaction Prediction","arxiv_id":"1704.08432","date":"2017-04-27","proceeding":null,"authors":["Sunyoung Kwon","Sungroh Yoon"],"abstract":"Chemical-chemical interaction (CCI) plays a key role in predicting candidate\ndrugs, toxicity, therapeutic effects, and biological functions. In various\ntypes of chemical analyses, computational approaches are often required due to\nthe amount of data that needs to be handled. The recent remarkable growth and\noutstanding performance of deep learning have attracted considerable research\nattention. However,even in state-of-the-art drug analysis methods, deep\nlearning continues to be used only as a classifier, although deep learning is\ncapable of not only simple classification but also automated feature\nextraction. In this paper, we propose the first end-to-end learning method for\nCCI, named DeepCCI. Hidden features are derived from a simplified molecular\ninput line entry system (SMILES), which is a string notation representing the\nchemical structure, instead of learning from crafted features. To discover\nhidden representations for the SMILES strings, we use convolutional neural\nnetworks (CNNs). To guarantee the commutative property for homogeneous\ninteraction, we apply model sharing and hidden representation merging\ntechniques. The performance of DeepCCI was compared with a plain deep\nclassifier and conventional machine learning methods. The proposed DeepCCI\nshowed the best performance in all seven evaluation metrics used. In addition,\nthe commutative property was experimentally validated. The automatically\nextracted features through end-to-end SMILES learning alleviates the\nsignificant efforts required for manual feature engineering. It is expected to\nimprove prediction performance, in drug analyses.","url_abs":"http://arxiv.org/abs/1704.08432v3","url_pdf":"http://arxiv.org/pdf/1704.08432v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deepcci-end-to-end-deep-learning-for-chemical","repo_url":"https://github.com/AstraZeneca/chemicalx","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"deepcci-end-to-end-deep-learning-for-chemical","repo_url":"https://github.com/cool21th/ai_drug_discovery","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"feature-engineering","task_name":"Feature Engineering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1704.08432","atlas_url":"https://app.syntology.ai/?focus=1704.08432","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}