Advanced Drug Delivery Reviews 151–152 (2019) 169–190 Contents lists available at ScienceDirect Advanced Drug Delivery Reviews journal homepage: www.elsevier.com/locate/addr The significance of artificial intelligence in drug delivery system design Parichehr Hassanzadeh ⁎, Fatemeh Atyabi, Rassoul Dinarvand Nanotechnology Research Center, Faculty of Pharmacy, Tehran University of Medical Sciences, Tehran 13169-43551, Iran article info abstract Article history: Over the last decade, increasing interest has been attracted towards the application of artificial intelligence (AI) Received 14 January 2019 technology for analyzing and interpreting the biological or genetic information, accelerated drug discovery, and Received in revised form 14 April 2019 identification of the selective small-molecule modulators or rare molecules and prediction of their behavior. Accepted 2 May 2019 Application of the automated workflows and databases for rapid analysis of the huge amounts of data and artifi- Available online 6 May 2019 cial neural networks (ANNs) for development of the novel hypotheses and treatment strategies, prediction of dis- fi fi Keywords: ease progression, and evaluation of the pharmacological pro les of drug candidates may signi cantly improve fi fi Artificial intelligence treatment outcomes. Target shing (TF) by rapid prediction or identi cation of the biological targets might be Artificial neural networks of great help for linking targets to the novel compounds. AI and TF methods in association with human expertise Target fishing may indeed revolutionize the current theranostic strategies, meanwhile, validation approaches are necessary to Drug delivery systems overcome the potential challenges and ensure higher accuracy. In this review, the significance of AI and TF in the development of drugs and delivery systems and the potential challenging issues have been highlighted. © 2019 Elsevier B.V. All rights reserved. Contents 1. Introduction.............................................................. 170 2. Artificialintelligence:thegeneralaspects................................................. 170 2.1. AIforbiomedicalapplications................................................... 172 2.2. TheroleofAIintheadvancementoftissueengineering....................................... 172 2.3. IntegrationofAIandnanotechnology............................................... 172 3. ApplicationofAIapproachesfordevelopmentofdrugsanddeliverysystems................................. 173 3.1. Peptidesynthesis........................................................ 174 3.2. Identifyingnovelantimycobacterialdrugs............................................. 174 3.3. Predictingtheeffectivenessofdrugdosinganddeliverymethods................................... 174 3.4. Rapid identificationofthebioactiveagentsandmonitoringofdrugrelease.............................. 174 3.5. Optimizingdrugreleasefrommatrixtablets............................................ 175 3.6. Making correlation between the formulation factors and release profiles............................... 176 3.6.1. Beads,pellets,andmicrospheres............................................. 177 3.6.2. Soliddispersions..................................................... 177 3.6.3. Implants........................................................ 177 3.6.4. Liposomes....................................................... 177 3.6.5. Transdermalformulations................................................ 178 3.6.6. Hydrodynamicallybalancedsystems,suppositories,pulmonarydelivery........................... 178 3.6.7. Controlled-releaseformulations.............................................. 178 3.6.8. Emulsions........................................................ 178 3.6.9. Microparticles...................................................... 178 3.6.10. Nanomaterials..................................................... 178 4. Applicationofnanorobotsfordrugdelivery................................................ 179 5. The significance of target fishing..................................................... 181 6. Challengingissuesandthepotentialsolutions............................................... 182 7. Conclusion............................................................... 183 ⁎ Corresponding author at: Nanotechnology Research Center, Faculty of Pharmacy, Tehran University of Medical Sciences, Fakhr Razi Ave, Tehran, Iran. E-mail addresses: [email protected] (P. Hassanzadeh), [email protected] (F. Atyabi), [email protected] (R. Dinarvand). https://doi.org/10.1016/j.addr.2019.05.001 0169-409X/© 2019 Elsevier B.V. All rights reserved. 170 P. Hassanzadeh et al. / Advanced Drug Delivery Reviews 151–152 (2019) 169–190 Roleofthefundingsource.......................................................... 183 Declarationofinterest............................................................ 184 Acknowledgements............................................................. 184 References.................................................................. 184 1. Introduction and continuous monitoring of glucose may significantly reduce the complications of diabetes. In this context, integration of glucose sensors, Over the last decades, development of the innovative systems for insulin delivery system, mathematical models and control algorithms is targeted delivery of therapeutics with maximal efficiency and minimal helpful. Integrating the insulin pump, dose calculator, and glucose side effects has attracted a growing interest [1–3]. Controlled drug de- meter into a device, an automated system has been provided for moni- livery and overcoming the challenges associated with conventional toring of glucose and delivery of insulin [3,21]. In designing intelligent drug delivery systems including the systemic toxicity, narrow therapeu- delivery systems, on-demand adjustment of dose or rate of drug release, tic index, and dose adjustment in long-term therapy have been the targeted delivery, and stability of pharmaceuticals should be taken into focus of intense research [2,3]. Application of the micro-fabrication account [22]. Regarding the self-monitoring delivery systems, appropri- technology for production of the implantable microchips appears prom- ate algorithms should be applied to control the amount and timing of ising for controlled delivery of drugs [4]. Microfabricated drug delivery drug release [3]. systems including the drug reservoirs with a variety of geometries or ca- Information technology, wireless communication, and artificial neu- pacities and capable of opening on command, provide continuous or ral networks (ANNs) contribute to the creation of smart drug delivery pulsatile drug delivery [5–7]. Using this type of multifunctional and so- systems that might be useful for overcoming the limitations of conven- phisticated devices for controlled drug release, a variety of challenges tional treatment strategies [2,3]. Wireless communication provides associated with traditional delivery systems have been addressed [8]. more flexibility for controlled drug delivery devices. The units receive For controlled and targeted delivery of therapeutics, application of the the instructions from the external sources, send data to the monitors, implantable drug delivery systems (Fig. 1) with ability of automatic and regulate drug release [3]. ANNs containing the interconnected pro- adjustment of drug concentration and timing of drug release is a prom- cessing elements which are created via simulating the network of ising approach for improving the efficiency, safety, and patients’ compli- model neurons, have been applied to develop software for mimicking ance [9]. This would be particularly important in chronic diseases which the biological processes, generating the control algorithms, pharmaco- require timely treatment and regular monitoring. dynamic/pharmacokinetic modeling, controlled drug delivery, and eval- Designing the implantable drug delivery systems necessitates con- uating the effectiveness of treatment strategies [3,23–29]. Using sideration of a number of points such as dose adjustment, targeted de- machine learning approaches for predicting ligand-based targets has re- livery, sustained release, and intelligent control system [9]. Neural sulted in the emergence of target fishing (TF) in which ligand datasets, networks, fuzzy logic, integrators, and differentiators have been applied target proteins, and relationship between the ligands and targets could for designing the control systems [10,11]. Methods of drug delivery in- be used to predict protein targets of novel compounds with biological clude the focused ultrasound, micro-pump mechanism, and targeted activities [28,29]. Indeed, application of high technologies is necessary delivery by microrobots [12–14]. For creation of the micro- or nanopar- for development of the next generations of drugs and innovative deliv- ticles for drug delivery, application of the microfluidic platforms is a ery systems. This review highlights the significance of artificial intelli- promising approach [15,16]. Using the microfluidic technology enables gence (AI) and TF in designing of drugs and delivery systems and the the development of smart drug delivery systems, e.g., Janus micro- or potential challenging issues. nanoparticles capable of delivery of multiple drugs [17,18]. For pro- grammed drug delivery, electronic components, wireless communica- fi tion hardware, and power supply have been embedded in a microchip 2. Arti cial intelligence: the general aspects implant