{"corpus_id":278955794,"paper_sha":"f2981a93e2af505c908077399144bbf2fb2cc274","doi":"10.1016/j.ejmech.2025.117825","arxiv_id":null,"pmid":40456205,"pmcid":null,"mag_id":null,"dblp_id":null,"acl_id":null,"title":"How generative Artificial Intelligence can transform drug discovery?","year":2025,"publication_date":"2025-05-01","venue":"European journal of medicinal chemistry","journal":{"name":"European journal of medicinal chemistry","pages":"\n          117825\n        ","volume":"295"},"journal_issn":null,"journal_title":null,"publication_types":["Review","JournalArticle"],"pubmed_pub_types":["Journal Article","Review"],"s2_fields_of_study":["Medicine","Computer Science"],"reference_count":147,"citation_count":8,"influential_citation_count":0,"is_open_access":false,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":[{"d":"Drug Discovery","mj":true,"ui":"D055808"},{"d":"Artificial Intelligence","mj":true,"ui":"D001185"},{"d":"Humans","mj":false,"ui":"D006801"},{"d":"Generative Artificial Intelligence","mj":false,"ui":"D000098842"}],"chemicals":null,"comments_corrections":null,"source_flags":5,"s2_open_access_pdf_url":null,"s2_open_access_landing_url":null,"s2_open_access_license":null,"s2_open_access_status":null,"pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Generative Artificial Intelligence (Generative AI) is transforming drug discovery by enabling advanced analysis of complex biological and chemical data. This review explores key Generative AI models, including Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), flow-based models and Transformer-based models, with Transformers gaining prominence due to the abundance of text-based biological data and the success of language models like ChatGPT. The paper discusses molecular representations, performance evaluation metrics, and current trends in Generative AI-driven drug discovery, such as protein-protein interactions (PPIs), drug-target interactions (DTIs) and de-novo drug design. However, these approaches face significant challenges, including applicability domain issues, lack of interpretability, data scarcity, novelty, scalability, computational resource limitations, and the absence of standardized evaluation metrics. These challenges hinder model performance, complicate decision-making, and limit the generation of novel and viable drug candidates. To address these issues, strategies such as hybrid models, integration of multiomics datasets, explainable AI (XAI) techniques, data augmentation, transfer learning, and cloud-based solutions are proposed. Additionally, a curated list of databases supporting drug discovery research is provided. The review concludes by emphasizing the need for optimized AI models, robust validation methods, interdisciplinary collaboration, and future academic efforts to fully realize the potential of Generative AI in advancing drug discovery.","claims":[{"public_id":"cl_131dd5c5e714c2644309f60cd8f9586b","status":"active","text":"Generative AI-driven drug discovery approaches face challenges that can hinder model performance, complicate decision-making, and limit the generation of novel and viable drug candidates.","confidence":0.9,"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/claims/cl_131dd5c5e714c2644309f60cd8f9586b"},{"public_id":"cl_67ddc711f94f0bba7ee70c3c2b82c49f","status":"active","text":"Hybrid models, integration of multiomics datasets, explainable AI techniques, data augmentation, transfer learning, and cloud-based solutions are proposed as strategies for addressing challenges in generative AI-driven drug discovery.","confidence":0.87,"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/claims/cl_67ddc711f94f0bba7ee70c3c2b82c49f"},{"public_id":"cl_797dd027cd2be076218f09485d3eb008","status":"active","text":"Optimized AI models, robust validation methods, and interdisciplinary collaboration are needed to more fully realize the potential of generative AI in drug discovery.","confidence":0.82,"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/claims/cl_797dd027cd2be076218f09485d3eb008"},{"public_id":"cl_ae3665061e6099325a8cfcce953c24e4","status":"active","text":"Transformer-based models are gaining prominence in generative AI-driven drug discovery because text-based biological data are abundant and language models such as ChatGPT have succeeded.","confidence":0.86,"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/claims/cl_ae3665061e6099325a8cfcce953c24e4"}],"concepts":[{"public_id":"co_148873aeb4453cb01f20a4169cac27dd","status":"active","name":"text-based biological data","description":"Biological information represented in textual form and used by language-model-based drug discovery approaches.","types":["data type"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_148873aeb4453cb01f20a4169cac27dd"},{"public_id":"co_231f970743ebbae3c94ddd9baf3302d3","status":"active","name":"drug discovery","description":"The biomedical research process of identifying and developing candidate therapeutic compounds.","types":["research area"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_231f970743ebbae3c94ddd9baf3302d3"},{"public_id":"co_452149ddf77885e9aeccabf8934d834b","status":"active","name":"hybrid models","description":"Modeling approaches that combine multiple AI methods or data sources to address drug discovery challenges.","types":["method"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_452149ddf77885e9aeccabf8934d834b"},{"public_id":"co_46464c270b3d7e203d869608c1b4de55","status":"active","name":"Generative Adversarial Networks","description":"A class of generative models considered for producing drug discovery-relevant outputs.","types":["model"],"aliases":["GANs"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_46464c270b3d7e203d869608c1b4de55"},{"public_id":"co_60e543ffdc0097a19e1f7483d00d4b89","status":"active","name":"multiomics datasets","description":"Integrated biological datasets spanning multiple omics data types for drug discovery modeling.","types":["dataset"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_60e543ffdc0097a19e1f7483d00d4b89"},{"public_id":"co_80a80e5241800a2d5726b3e01bd86cb5","status":"active","name":"Variational Autoencoders","description":"A class of latent-variable generative models considered for molecular and drug discovery tasks.","types":["model"],"aliases":["VAEs"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_80a80e5241800a2d5726b3e01bd86cb5"},{"public_id":"co_9a133830ee5f4116f6c3b9ecb90feedc","status":"active","name":"explainable AI techniques","description":"Methods intended to make AI model behavior more interpretable in drug discovery.","types":["method"],"aliases":["XAI techniques"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_9a133830ee5f4116f6c3b9ecb90feedc"},{"public_id":"co_9bb703f022dc0be4c81561522cb86aa8","status":"active","name":"standardized evaluation metrics","description":"Common measurement criteria for assessing generative AI models in drug discovery.","types":["evaluation metric"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_9bb703f022dc0be4c81561522cb86aa8"},{"public_id":"co_b0c5f1dfd21169a8311c852baed59f2d","status":"active","name":"Transformer-based models","description":"Neural network models based on Transformer architectures and applied to text-rich biological data.","types":["model"],"aliases":["Transformers"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_b0c5f1dfd21169a8311c852baed59f2d"},{"public_id":"co_c702bf40e076e37e6a571173ecb7dc32","status":"active","name":"drug-target interactions","description":"Relationships between drugs and biological targets considered in AI-driven drug discovery.","types":["biological relationship"],"aliases":["DTIs"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_c702bf40e076e37e6a571173ecb7dc32"},{"public_id":"co_d271f6c39a3771f5f59f5e8f8fb40370","status":"active","name":"robust validation methods","description":"Validation approaches intended to reliably assess AI-generated drug discovery outputs.","types":["evaluation method"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_d271f6c39a3771f5f59f5e8f8fb40370"},{"public_id":"co_d913468fbd91914e08ac612a5180f2d0","status":"active","name":"protein-protein interactions","description":"Biological relationships between proteins considered as a drug discovery application area.","types":["biological relationship"],"aliases":["PPIs"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_d913468fbd91914e08ac612a5180f2d0"},{"public_id":"co_e7f8c2e1428aeede9c60ccb212a4fc83","status":"active","name":"Generative Artificial Intelligence","description":"An artificial intelligence approach used here to generate and analyze biological and chemical data for drug discovery.","types":["technology"],"aliases":["Generative AI"],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_e7f8c2e1428aeede9c60ccb212a4fc83"},{"public_id":"co_edc8ef3ef55668d3cb08544f939f5195","status":"active","name":"flow-based models","description":"Generative models that transform probability distributions for data generation in drug discovery contexts.","types":["model"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_edc8ef3ef55668d3cb08544f939f5195"},{"public_id":"co_fc3fdec90d62b141b1847a05c448d588","status":"active","name":"de-novo drug design","description":"The task of designing new drug candidates rather than selecting only from existing compounds.","types":["application"],"aliases":[],"contributors":[{"id":35,"public_id":"b2adb6bfad","public_label":"Anonymous (b2adb6bfad)","roles":["extraction"],"url":"https://sah.borca.ai/u/b2adb6bfad"},{"id":2,"public_id":"4715169a40","public_label":"AK (4715169a40)","roles":["review"],"url":"https://sah.borca.ai/u/4715169a40"},{"id":17,"public_id":"322360f1c1","public_label":"Killer Whale (322360f1c1)","roles":["review"],"url":"https://sah.borca.ai/u/322360f1c1"}],"url":"https://sah.borca.ai/concepts/co_fc3fdec90d62b141b1847a05c448d588"}],"external_ids":{"DOI":"10.1016/j.ejmech.2025.117825","ArXiv":null,"PubMed":40456205,"PubMedCentral":null,"MAG":null,"DBLP":null,"ACL":null},"open_access":{"is_open_access":false,"pdf_url":null,"landing_url":"https://sah.borca.ai/papers/278955794","source":null,"pdf_url_source":null,"license":null,"reason":"pdf_url_not_indexed"},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":5},"paper_id":631908,"paper_uid":"92589138-4899-48aa-b591-1888054cfd43","canonical_identity":{"paper_id":631908,"paper_uid":"92589138-4899-48aa-b591-1888054cfd43","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/278955794"}