Artificial Intelligence Meets Drug Discovery: A Systematic Review on AI-Powered Target Identification and Molecular Design
Notice bibliographique
Résumé
Background: Drug discovery is an inherently complex, resource-intensive, and time-consuming process, often requiring more than a decade to progress from initial target identification to regulatory approval. Despite technological advancements, high attrition rates and escalating research costs remain significant barriers. Artificial intelligence (AI) has emerged as a transformative force in pharmaceutical research, integrating machine learning, deep learning, and computational biology to revolutionize drug discovery processes. AI-driven models enable faster target identification, molecular docking, lead optimization, and drug repurposing, offering unprecedented efficiency in discovering novel therapeutics. Objective: This systematic review aims to assess the integration of AI in drug discovery, focusing on its applications in target identification, structure-based drug design (SBDD), ligand-based drug design (LBDD), and predictive modeling for clinical trials. The review highlights AI-driven advancements in molecular simulations, de novo drug design, and biomarker identification, providing insights into how AI enhances pharmaceutical innovation. Methods: A comprehensive literature review was conducted using PubMed, Scopus, Web of Science, and Google Scholar, covering research published between 2015 and 2025. Studies evaluating AI applications in computational drug discovery, virtual screening, molecular dynamics, and predictive analytics were included. The methodologies analyzed include convolutional neural networks (CNNs), generative adversarial networks (GANs), reinforcement learning models, and graph neural networks (GNNs). Quality assessment was performed using the Newcastle-Ottawa Scale and the Cochrane Risk of Bias Tool to ensure the reliability of selected studies. Results: AI-powered approaches have significantly improved drug discovery by enhancing target identification, optimizing molecular docking simulations, and accelerating lead optimization. Notable achievements include DeepMind’s AlphaFold in protein structure prediction, Insilico Medicine’s AI-driven fibrosis drug (which entered clinical trials in under 18 months), and BenevolentAI’s identification of Janus kinase inhibitors (JAK) for COVID-19 treatment. AI has also been instrumental in drug repurposing efforts, uncovering new therapeutic potentials for existing FDA-approved drugs. Furthermore, AI-enhanced clinical trial modeling and patient stratification have improved trial efficiency and success rates. Challenges & Future Directions: Despite its potential, AI-driven drug discovery faces challenges, including data bias, lack of interpretability, regulatory barriers, and ethical concerns regarding AI-generated predictions. To maximize AI’s impact, future research should focus on standardizing biological datasets, integrating multi-omics data, and developing explainable AI (XAI) models. The emergence of quantum AI and hybrid AI-physics models presents promising avenues for further accelerating drug discovery. Conclusion: AI is reshaping the landscape of drug discovery, offering unparalleled efficiency in identifying novel drug candidates, optimizing molecular interactions, and predicting clinical outcomes. Its integration with biological data and computational simulations paves the way for the development of personalized and highly effective therapeutics. However, addressing AI-related challenges in transparency, validation, and regulatory compliance remains crucial for translating AI-generated discoveries into clinically viable treatments.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction distillée sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Apprise à partir de 10 348 étiquettes directes de Codex et de 10 348 étiquettes directes de Gemma. Le mode candidate est l'union des têtes enseignantes seuillées; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont ni des étiquettes humaines ni des étiquettes directes de modèles de pointe.
Scores Codex et Gemma par catégorie
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,006 | 0,004 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,001 |
| Méta-épidémiologie (sens large) | 0,002 | 0,001 |
| Bibliométrie | 0,000 | 0,001 |
| Études des sciences et des technologies | 0,000 | 0,000 |
| Communication savante | 0,000 | 0,001 |
| Science ouverte | 0,002 | 0,001 |
| Intégrité de la recherche | 0,000 | 0,001 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,000 | 0,001 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule tête enseignante, pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».