Perspective on the challenges and opportunities of accelerating drug discovery with artificial intelligence

Maria, John P. Santa; Wang, Yuan; Camargo, Luiz Miguel

doi:10.3389/fbinf.2023.1121591

Cited by 10 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In case of drug discovery, it is challenging to get well curated data when it comes about the rare diseases or novel targets (Dong et al, 2022). Furthermore, there may be the chances of overfitting in case of machine learning model, limitations in the chemical space exploration since the novel drug candidates or unconventional chemical structures may not be well captured by the trained AI/ML models (Gysi et al, 2023).…”

Section: Discussionmentioning

confidence: 99%

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal

Priyadarsinee,

Jamir,

Nagamani

et al. 2023

Preprint

View full text Add to dashboard Cite

Computational drug discovery is intrinsically interdisciplinary and has to deal with the multifarious factors which are often dependent on the type of disease. Molecular Property Diagnostic Suite (MPDS) is a Galaxy based web portal which was conceived and developed as a disease specific web portal, originally developed for tuberculosis (MPDSTB). As specific computational tools are often required for a given disease, developing a disease specific web portal is highly desirable. This paper emphasises on the development of the customised web portal for COVID-19 infection and is referred to as MPDSCOVID-19. Expectedly, the MPDS suites of programs have modules which are essentially independent of a given disease, whereas some modules are specific to a particular disease. In the MPDSCOVID-19 portal, there are modules which are specific to COVID-19, and these are clubbed in SARS-COV-2 disease library. Further, the new additions and/or significant improvements were made to the disease independent modules, besides the addition of tools from galaxy toolshed. This manuscript provides a latest update on the disease independent modules of MPDS after almost 6 years, as well as provide the contemporary information and tool-shed necessary to engage in the drug discovery research of COVID-19. The disease independent modules include file format converter and descriptor calculation under the data processing module; QSAR, pharmacophore, scaffold analysis, active site analysis, docking, screening, drug repurposing tool, virtual screening, visualisation, sequence alignment, phylogenetic analysis under the data analysis module; and various machine learning packages, algorithms and in-house developed machine learning antiviral prediction model are available. The MPDS suite of programs are expected to bring a paradigm shift in computational drug discovery, especially in the academic community, guided through a transparent and open innovation approach. The MPDSCOVID-19 can be accessed at http://mpds.neist.res.in:8085.

show abstract

Section: Discussionmentioning

confidence: 99%

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal

Priyadarsinee,

Jamir,

Nagamani

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Most of the well-known success stories of AI have been in image recognition (e.g., in the early days, the approach was trained to for instance recognize cat and dog images, but today the method can be used to analyze biopsies or guide surgery) while also advertised in reducing time to reach phase I clinical trial. In the latter case, one can site the story of compound DSP-1181, developed by Exscientia and Sumitomo Dainippon Pharma, intended to treat obsessive compulsive disorder where time from first screening to the development stage was 4 time faster than using a conventional approach (although, unfortunately, the molecule failed in phase I, for numerous reasons including a difficult target while it was also observed that the molecules generated by AI were not novel) (Santa Maria Jr et al, 2023) (https://www.science.org/content/blog-post/another-aigenerated-drug; https://www.cas.org/resources/cas-insights/drugdiscovery/ai-designed-drug-candidates). Similar observations have been posted by hundreds of financial analysts and research scientists about results obtained by other AI companies.…”

Section: Artificial Intelligence: Trust But Verifymentioning

confidence: 99%

Drug discovery and development: introduction to the general public and patient groups

Singh

Vayer

Tanwar

et al. 2023

Front. Drug Discov.

View full text Add to dashboard Cite

Finding new drugs usually consists of five main stages: 1) a pre-discovery stage in which basic research is performed to try to understand the mechanisms leading to diseases and propose possible targets (e.g., proteins); 2) the drug discovery stage, during which scientists search for molecules (two main large families, small molecules and biologics) or other therapeutic strategies that interfere or cure the investigated disease or at least alleviate the symptoms; 3) the preclinical development stage that focuses on clarifying the mode of action of the drug candidates, investigates potential toxicity, validates efficacy on various in vitro and in vivo models, and starts evaluate formulation; 4) the clinical stage that investigates the drug candidate in humans; 5) the reviewing, approval and post-market monitoring stage during which the drug is approved or not. In practice, finding new treatments is very challenging. Despite advances in the understanding of biological systems and the development of cutting-edge technologies, the process is still long, costly with a high attrition rate. New approaches, such as artificial intelligence and novel in vitro technologies, are being used in an attempt to rationalize R&D and bring new drugs to patients faster, but several obstacles remain. Our hope is that one day, it becomes possible to rapidly design inexpensive, more specific, more effective, non-toxic, and personalized drugs. This is a goal towards which all authors of this article have devoted most of their careers.

show abstract

“…With AI's ability to accelerate every step of drug development, the possibility of realizing this goal in the future becomes potentially plausible. [43][44][45][46] Additionally, AI and LLMs could also be used to predict the potential misuse potential of certain compounds or their precursors for harmful purposes. By identifying these 'red flag' compounds, it would be possible to implement preventive measures, such as increased oversight or restrictions on their use.…”

Section: Countermeasures and Antidotesmentioning

confidence: 99%

“…Nevertheless, it is crucial to be mindful of the rapid evolution of AI in various stages of drug development which could potentially expedite the process of registering such antidotes for clinical use. With AI's ability to accelerate every step of drug development, the possibility of realizing this goal in the future becomes potentially plausible 43–46 …”

Section: Recommendationsmentioning

confidence: 99%

Artificial intelligence in clinical pharmacology: A case study and scoping review of large language models and bioweapon potential

Rubinic,

Kurtov,

Rubinic

et al. 2023

Brit J Clinical Pharma

View full text Add to dashboard Cite

AimThis paper aims to explore the possibility of employing large language models (LLMs) – a type of artificial intelligence (AI) – in clinical pharmacology, with a focus on its possible misuse in bioweapon development. Additionally, ethical considerations, legislation, and potential risk reduction measures are analysed.MethodsA comprehensive review of existing literature investigates the potential misuse of AI and LLMs in bioweapon creation. The search includes articles from PubMed, Scopus, and Web of Science Core Collection that were identified using a specific protocol. To explore the regulatory landscape, the OECD.ai platform was used.Resultshe review highlights the dual‐use vulnerability of AI and LLMs, with a focus on bioweapon development. Subsequently, a case study is used to illustrate the potential of AI manipulation resulting in harmful substance synthesis. Existing regulations inadequately address the ethical concerns tied to AI and LLMs. Mitigation measures are proposed, including technical solutions (explainable AI), establishing ethical guidelines through collaborative efforts, and implementing policy changes to create a comprehensive regulatory framework.ConclusionThe integration of AI and LLMs into clinical pharmacology presents invaluable opportunities, while also introducing significant ethical and safety considerations. Addressing the dual‐use nature of AI requires robust regulations, as well as adopting a strategic approach grounded in technical solutions and ethical values following the principles of transparency, accountability, and safety. Additionally, AI’s potential role in developing countermeasures against novel hazardous substances is underscored. By adopting a proactive approach, the potential benefits of AI and LLMs can be fully harnessed while minimizing the associated risks.

show abstract

Perspective on the challenges and opportunities of accelerating drug discovery with artificial intelligence

Cited by 10 publications

References 30 publications

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal

Drug discovery and development: introduction to the general public and patient groups

Artificial intelligence in clinical pharmacology: A case study and scoping review of large language models and bioweapon potential

Contact Info

Product

Resources

About

Perspective on the challenges and opportunities of accelerating drug discovery with artificial intelligence

Cited by 10 publications

References 30 publications

Molecular Property Diagnostic Suite for COVID-19 (MPDSCOVID-19): An open access disease specific drug discovery portal

Molecular Property Diagnostic Suite for COVID-19 (MPDSCOVID-19): An open access disease specific drug discovery portal

Drug discovery and development: introduction to the general public and patient groups

Artificial intelligence in clinical pharmacology: A case study and scoping review of large language models and bioweapon potential

Contact Info

Product

Resources

About

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal

Molecular Property Diagnostic Suite for COVID-19 (MPDS^COVID-19): An open access disease specific drug discovery portal