Impact of Artificial Intelligence and ChatGPT on Pharmaceutical Industries

Nilesh R. Suryawanshi, Karan. A. Patil, Nikhil. J. Rajput, H. P. Suryawanshi*,

R. A. Ahirrao,   J. I. Pinjari

P. G. College of Pharmaceutical Science and Research, Chaupale, Nandurbar, (MS) India. Correspondence: hemant.surya@gmail.com Mob: 9923032073

INTRODUCTION

The use of artificial intelligence in pharmaceutical technology has increased over time, and it can help save time and money while also improving our understanding of the connections between various formulations and process variables. It might be regarded as the first AI programmer; it was developed by Newell and Simon in 1995.John McCarthy is known as the father of artificial intelligence and the person who finally invented the term. The ability of a digital computer or computer-controlled robot to carry out actions frequently associated with intelligent beings is known as artificial intelligence (AI).

When was AI introduced?

Classical thinkers' attempts to define human thought as a symbolic system can be linked to the birth of AI in 1956. But the phrase "artificial intelligence" (AI) wasn't actually coined until a conference at Dartmouth College in Hanover, New Hampshire, in 1956.

The birth of AI (1952-1956) –

Ø  Allen Newell and Herbert A. Simon developed the "Logic theorist" system, the first artificial intelligence system, in 1955. This programmer proved 38 out of 52 mathematical theorems in addition to finding new and better proofs for some of the theorems.

Ø  John McCarthy, a computer scientist from the US, developed the term "AI" in 1956 at the Dartmouth conference. Initially, academics embraced AI as an actual topic for research [1, 2].

Agents, compounds, genes, and other relevant knowledge a specific to a given disease can be recognized by ChatGPT. ChatGPT can let researchers from various fields talk to one another and offer essential details for their development of novel drugs. ChatGPT can create new structures for molecules based on their chemical and physical characteristics, aiding researchers in more successful drug creation. Furthermore, ChatGPT can forecast a compound's pharmacokinetic (PK), pharmacodynamics (PD), and toxic features, ranging providing crucial information for drug development [3].

2. AI in Pharmaceutical Industry

An enormous amount of estimated US$ 2.6 billion over more than 10 years is spent on developing a drug. In spite of this spending nine out of ten candidate therapies fail between phases I trials and regulatory approval.{4} (Fleming, 2018) Artificial intelligence in pharmaceuticals refers to the a job of automated algorithms to carry out tasks that normally need human intellect, as stated in the report by Coding Arsine, CEO of Digital Authority Partners. The application of artificial intelligence in the pharmaceutical and biotech industry has altered how scientists create fresh drugs to treat ailments, and more over the last five years [5].

Leading pharma companies are progressing towards using AI for certain strategic applications. AI will help the pharma companies to increase the success rates of new drug discovery meanwhile decreasing operational costs [6]. With the help of big data, AI, and machine learning, McKinsey projects that enhanced decision-making, optimized innovation, improved testing performance, clinical trials, and the development of new tools could generate up to $100 billion in pharmaceuticals and health care goods each year [7].

4. Application of AI and ChatGPT in the Pharmaceutical Industry:

AI mainly Applicable For pharmaceutical industries for:

1.      Drug discovery.

2.      Clinical Trials.

3.      Drug development.

4.      Disease prevention.

5.      Manufacturing.

6.      QA/QC.   

7.      Novel medication and drug delivery systems.

8.      Drug discovery:

The vast chemical space, comprising >10⁶⁰ molecules, fosters the development of a large number of drug molecules [8]. However, the medication development process is constrained by a lack of advanced technology, making it a time-consuming and expensive try that can be resolved by using AI [9]. AI can identify hit and lead compounds, expedite drug target validation, and improve drug structure design [10].Figure 1 shows a number of ways AI is employed to find new drugs.

Figure 1: Applications of artificial intelligence (AI) in different subfields of the pharmaceutical industry [11]

Despite its benefits, AI still has a number of key data hurdles to overcome, including the size, expansion, diversity, and predictability of the data. Millions of molecules may be present in the data sets available to pharmaceutical companies for medication development, making it likely that typical ML systems cannot handle this kind of data. Large numbers of chemicals or basic physicochemical properties, such as log P or log D, can be predicted quickly using a computational model based on the quantitative structure-activity relationship (QSAR). However, these models are far from being able to anticipate complicated biological features like a compound's effectiveness and unfavorable side effects. Furthermore, QSAR-based models encounter issues such as short training sets and experimental data inaccuracies in [12, 13].

Figure 2: Role of artificial intelligence (AI) in drug discovery [14]

AI is used to create unique chemicals with specific features and activities. The identification and change of already-existing molecules is a common stay of conventional approaches, but it may be a labor- and time-intensive method. On the other hand, AI-based techniques can facilitate the quick and effective design of new compounds with desirable characteristics and activities. To suggest new medicinal molecules, for instance, a deep learning (DL) system has recently been trained on a collection of known medicinal compounds and their accompanying attributes [15].

4.1 AI has revolutionized drug research and discovery in numerous ways: [16]

Some of the key contributions of AI in this domain include the following:

1. Target Identification: To find potential therapy targets, AI systems may evaluate a variety of data sources, such as genomic, proteomic, and clinical data. AI aids in the creation of drugs that can modify biological processes by identifying disease-associated targets and molecular pathways.

2. Virtual Screening: Through the effective use of AI, large chemical libraries can be quickly tested to find drug candidates that have a strong chance of attaching to a particular target. AI assists scientists in prioritizing and picking compounds for experimental testing, saving time and resources by simulating chemical interactions and figuring out binding affinities.

3. Structure-Activity Relationship (SAR) Modeling: The chemical structure of substances and their biological function can be connected using AI models. By creating compounds with desirable characteristics, such as high potency, selectivity, and favorable pharmacokinetic profiles, researchers can optimize therapeutic prospects.

4. De Novo Drug Design: AI algorithms can suggest novel drug-like substances by using reinforcement learning and generative models. AI broadens the chemical landscape and assists in the creation of novel drug candidates by learning from chemical libraries and experimental data.

5. Optimization of Drug Candidates: A variety of parameters, such as efficacy, safety, and pharmacokinetics, can be taken into account by AI algorithms when analyzing and optimizing drug candidates. This aids scientists in modifying drugs to increase their efficacy while reducing potential negative effects. Repurposing of Drugs Large-scale biomedical data can be analyzed using AI approaches to find existing medications with potential therapeutic value for various ailments. AI speeds up and reduces the cost of the drug discovery process by repurposing current drugs for new applications.

6. Toxicity Prediction: By examining the chemical makeup and properties of substances, AI systems can forecast drug toxicity. Machine learning algorithms that have been trained on toxicological databases can predict negative effects or recognize potentially dangerous structural characteristics. This aids in the selecting of safer compounds and the prevention of potential negative effects in clinical trials.

4.2Artificial intelligence tools used in drug discovery:

Table 1: Examples of AI tools used in drug discovery [17]

4.3 ChatGPT in drug discovery:

The ChatGPT model for languages was created by OpenAI. Because it was trained on a big dataset of human language, the machine learning model can produce text that resembles that of a human. It can be used for a variety of NLP tasks, including as text summarization, language translation, and answering question. In our recent work, we looked at the potential use of ChatGPT to drug discovery [18].

OpenAI launched Chat GPT, a modern language model. Given that it was trained on a sizable text corpus, this deep neural network can respond to various cues in a way that is human-like. When developing languages, Chat GPT may handle context and ongoing dependencies using the transformer design. It may be trained to carry out particular tasks like language translation, question answering, and text completion and produce text that resembles human speech for a range of objectives [19].

The process of finding and creating new drugs to treat ailments is known as drug discovery. A number of procedures, spanning target selection, lead discovery, preclinical development, clinical trials, and regulatory approval, participate in this multidisciplinary, complex discipline. Finding safe and efficient treatments for a variety of ailments is the aim of drug discovery. The use of computer-based techniques to aid in the creation of new medical products is referred to as "drug discovery using computational chemistry20, 21, 22].

4.4.1Future prospects of ChatGPT in drug discovery:

1. Identifying and validating novel drug targets: ChatGPT can be adjusted using a dataset of scientific literature to produce summaries of the most recent findings on a particular disease or biological target. As a result, researchers may be able to find new potential targets faster or understand the current state of research in an area better.

 2. Developing new chemical compounds with comparable properties enables: ChatGPT to be optimized on a dataset of known drug-like chemicals. Researchers may be able to find new lead molecules this way that have a greater chance of working in both preclinical and clinical tests.

3. Improving drug properties: ChatGPT can be used to aid in the virtual screening of chemical libraries during the early stages of drug discovery and to forecast the pharmacokinetics and pharmacodynamics of novel medications.

4. Assessing toxicity: ChatGPT can be fine-tuned on a dataset of toxicity data and used to predict the potential toxic effects of new drugs [23].

2. Clinical Tials:

Clinical trials take about 6-7 years to finish and include a substantial monetary investment in order to find out the efficacy and safety of a medicinal product in people for a specific disease condition. Only one out of every ten molecules that go through these trials, however, obtains successful clearance, which represents a significant loss for the industry [24]. These mistakes could be brought on by poor patient selection, a dearth of technological needs, or weak infrastructure. Due to the volume of digital medical data that is already available, these issues can be reduced with the use of AI. The patient enrollment phase of a clinical study takes up one-third of the total time spent. Selecting the right subjects can stop 86% of clinical trials from failing that would otherwise take place. AI can assist in the selection of only a certain diseased group for recruitment in Phase II and III of clinical trials by using patient-specific genome-exposure profile analysis. This can aid in the quick recognition of the therapeutic targets that would be optimal for the selected patients [25].

3. Drug development:

The practical usage of AI has the potential to boost R&D. AI is capable of a thing, from developing and detecting novel substances to researching and validating target-based medicine [26]. The identification, optimization, and commercialization of new medications are being revolutionized by this fusion of cutting-edge technology and biomedical research. The use of AI in drug discovery has the potential to significantly speed up the development of new treatments, improve the efficacy of existing ones, and open the door to more individualized medical care.

Therapeutic discovery has historically been a labor- and resource-intensive process that entails selecting possible therapeutic candidates, evaluating their efficacy and safety, and overcoming regulatory barriers to find disease-related targets [27].

Artificial intelligence (AI) has become an important tool in drug discovery and development. These technologies can be used to analyses large datasets, identify patterns and relationships, and make predictions about potential drug targets and candidate compounds [28].A critical phase in the development of novel medications is the discovery of hit-and-lead compounds. ChatGPT can let researchers from different disciplines communicate with one another and offer vital information for the creation of new drugs. ChatGPT can create new structures for molecules depending on their physical and chemical properties, aiding researchers in more successful drug development. Additionally, ChatGPT can forecast a compound's pharmacokinetic (PK), pharmacodynamics (PD), and toxicity characteristics, offering crucial information for drug development [29,30].

4. Disease prevention:

Drug companies can employ AI to create cures for both common diseases like Parkinson's and Alzheimer's as well as unusual disorders. Since the ROI is so poor compared to the time and money required to create drugs to treat uncommon conditions, pharmaceutical companies typically do not devote their time and resources to searching for cures for rare diseases [31]. Using AI, big pharma is finding faster and more reliable patients for clinical trials, introducing automated robot pharmacies to fill prescriptions and dispense pharmaceuticals, and increasing marketing, logistics, and supply chain. AI is also assisting big pharma with drug adherence and dosage. The future of pharma is AI, yet the technology is already available. Costs can be reduced, new, efficient treatments can be developed and above all, artificial intelligence can help save lives. Hence, biotech companies should begin utilizing the advantages of AI as soon as possible. [32]

5. Manufacturing:

Modern manufacturing systems are attempting to impart human knowledge to machines as a result of the growing complexity of manufacturing processes as well as an increasing need for efficiency and greater product quality. This is continuously changing the production process [33]. It has been successfully used for the synthesis and manufacture of sildenafil, diphenhydramine hydrochloride, and rufinamide, with the yield and purity significantly similar to manual synthesis [34].AI technologies have the potential to efficiently complete granulation in granulators with capacities ranging from 25 to 600 [35].

In order to anticipate the amount of granulation fluid to be added, the necessary speed, and the diameter of the impeller in both geometrically identical and dissimilar granulators, the researchers produced a polynomial equation [36]. In order to decrease tablet capping on the production line, ANNs and fuzzy models investigated the relationship between machine settings and the capping issue [37]. The application for a patent describes a system that employs a processor that receives patient data to determine the ideal drug and dose regimen for each patient before manufacturing the appropriate transdermal patch [38].

6. QA/QC:

A variety of elements must be controlled during the manufacturing process of the desired product from raw materials [39]. By applying a "Quality by Design" approach, the FDA modified Current Good Manufacturing Practices (cGMP) in order to better understand the crucial process and precise standards that determine the ultimate quality of the pharmaceutical product. AI can also be used to control in-line manufacturing processes so that the product meets the standards that are desired [40].With the use of ANN, Go et al. examined the dissolution profile, a sign of the consistency of theophylline pellets from batch to batch. With an error of only 8%, the predicted dissolution of the formulation is correct [41].

7. Novel medication drug delivery systems:          

Drug distribution procedures may shift in a number of ways thanks to AI. AI can be used, for instance, to identify new therapeutic targets, improve medication formulations, and forecast the efficacy of treatments and their toxicity. The personalization of medicine is a promising area where AI can be used in drug delivery systems.

5. The Role of Collaboration between AI Researchers and Pharmaceutical Scientists:

In order to create new and effective treatments for a variety of ailments, coordination between AI researchers and pharmaceutical scientists is essential. They can produce potent algorithms and machine-learning models that predict the efficacy of possible drug candidates and expedite the drug development process by combining their knowledge and skills. As AI algorithms may be used to assess the data gathered during clinical trials to discover trends and the potential side effects of the pharmaceuticals being tested, this partnership can also help improve the accuracy and efficiency of clinical studies. This can expedite the entire drug development process and assist pharmaceutical corporations in making educated judgments about which drug candidates to pursue. A case in point is the partnership between the pharmaceutical. [42]

Various pharmaceutical companies have invested in and are continuing to invest in AI and have collaborated with AI companies to develop essential healthcare tools. The collaboration of Deep Mind Technologies, a subsidiary of Google, with the Royal Free London NHS Foundation Trust for the assistance of acute kidney injury is an example of this. Major pharmaceutical companies and AI players are detailed in Figure 3. [43]

Figure 3: Leading pharmaceutical companies and their association with Artificial Intelligence (AI)

CONCLUSION

The current healthcare sector is facing several complex challenges, such as the increased cost of drugs and therapies, and society needs specific significant changes in this area. With the inclusion of AI in the manufacturing of pharmaceutical products, personalized medications with the desired dose, release parameters, and other required aspects can be manufactured according to individual patient need.

ACKNOWLEDGEMENT

We are thankful to the Principal and management of P. G. College of Pharmaceutical Science and Research Chaupale, Nandurbar, for providing moral support and necessary facilities during complete of this work.

REFERENCES       

1.      Vyas M., T. S. Asian J Pharmaceutics, [2018]72-76.

2.      M, M.Research gate,[2015].

3.      Davies, N. M. Adapting artificial intelligence into the evolution of pharmaceutical sciences and publishing: Technological darwinism. Pharm Science.[2023]

4.      Fleming, N. How artificial intelligence is changing drug discovery,[2018].

5.      Digital Authority Partners. (2019 November 23). Retrieved from https://www.digitalauthority.me/resources/artificial-intelligence-pharma/

6.      Berg, K. M. (2023, November 23). Using Artificial Intelligence for Drug Discovery. Retrieved from http://www.medcitynews.

7.      Cattell J, C. S. (2013, April). How big data can revolutionize pharmaceutical R and D. Retrieved from https://www.mckinsey.com/industries/life-sciences/our-insights/how-big-data-can-revolutionize-pharmaceutical-r-and-d

8.      Kit-Kay Mak, M. R. Artificial intelligence in drug development: present status and future prospects. Drug Discovery Today [2019].

9.      M.et al, V. Artificial intelligence:the beginning of a new era in pharmacy. Scopus, [2018]72-76.

10.  Matthew A Sellwood1, M. A. Artificial intelligence in drug discovery. Future Med. Chem, [2018]

11.  Artificial intelligence in drug. (2021). Retrieved from https://www.bing.com/images/search?view=detailv2&insightstoken=bcid_qCoBHe9jOBwG1g*ccid_KgEd72M4*cp_DB17948D748D1E5BEBC343CA9E75D2E6&form=WCVSID&iss=SBIUPLOADGET&darkschemeovr=1&selectedindex=4&id=614416574&ccid=KgEd72M4&exph=270&expw=836&vt=2&simid=60798

12.  Zhu, H. Big Data and Artificial Intelligence Modeling for Drug Discovery. Annual Review of Pharmacology and Toxicology , [2020] 573-589.

13.  Ciallella, H. a. Advancing Computational Toxicology in the Big Data Era by Artificial Intelligence: Data-Driven and Mechanism-Driven Modeling for Chemical Toxicity. Chem. Res. Toxicol., [2019] 536–547.

14.  Role of artificial intelligence (AI) in drug discovery.[2021]Retrieved from https://www.bing.com/images/search?view=detailv2&insightstoken=bcid_qHAnI9-kLRwGdw*ccid_cCcj36Qt*cp_3B33FD597D5DBFC04F8203D108B11C9B&form=WCVSID&iss=SBIUPLOADGET&darkschemeovr=1&selectedindex=0&id=-313176489&ccid=cCcj36Qt&exph=351&expw=740&vt=2&simid=6080

15.  Automatic Chemical Design Using a Data-Driven Continuous. ACS Central Sci., [2018] 268–276.

16.  Valavanidis, P. A. Artificial Intelligence (AI) Applications.Τhe most important technology we ever develop and we must ensure it . Research Gate.[2023].

17.  Debleena Paul, G. S. Artificial intelligence in drug. Drug Discovery Today.[2021]

18.  Aydın, Ö., & Karaarslan, E. J. OpenAI ChatGPT generated literature review: Digital [2022].

19.  Jiao, W., Wang, W., Huang, J.-t., Wang, X., & Tu, Z. J. Is ChatGPT a good translator? A preliminary study[2023].

20.  Gangadevi, S., Badavath, V. N., Thakur, A., Yin, N., De Jonghe, S., Acevedo, O., . . . Neyts, J. J. Kobophenol A inhibits binding of host ACE2 receptor with spike RBD domain of SARS-CoV-2, a lead compound for blocking COVID-19. [2021]12 (7), 1793-1802.

21.  Madhyastha, H., Madhyastha, R., Thakur, A., Kentaro, S., Dev, A., Singh, S., . . . Biointerfaces, S. B. c-Phycocyanin primed silver nano conjugates: Studies on red blood cell stress resilience mechanism. [2020] 194, 111211.

22.  Sharma, G., Song, L. F., & Merz, K. M. Modeling, Effect of an Inhibitor on the ACE2-Receptor-Binding Domain of SARS-CoV-2.[2022]

23.  Thakur, A., Mandal, S. C., Banerjee, S. J., & Studies, M. Compounds of natural origin and acupuncture for the treatment of diseases caused by estrogen deficiency. [2016] 109-117.

24.  Hay, M. e. Clinical development success rates for investigational drugs. Nat.Biotechnology, [2014] 40-51.

25.  Harrer, S. e. Artificial intelligence for clinical trial design. . Trends Pharmacological science, [2019]577-551.

26.  Kamal H., L. V. Front Neurol. [2018]

27.  Tariq, A. G.Evaluating the Potential of Artificial Intelligence in Orthopedic Surgery for Value-based Healthcare. . International Journal of Multidisciplinary Sciences and Arts, [2023] 27-36.

28.  Mak, K.-K. a. Artificial intelligence in drug development: present status and future prospects. Drug Discovery Today, [2019]773–780.

29.  Asselah, T. D. COVID-19:Discovery, diagnostics and drug development. J. Hepatol, [2021] 168–184.

30.  Azodi, C. B. Opening the black box: Interpretable machine learning for geneticists. Trends Genet, [2020] 442–455.

31.  Feng R., B. M. (2018). 358-362.

32.  Bajorath J, K. S.  The Future Is Now: Artificial Intelligence in Drug Discovery. Journal of Medicinal Chemistry,[2019] 5249-5249. .

33.  Meziane, F. e. Intelligent systems in manufacturing: current developments and future prospects. [2000] 218-238.

34.  Steiner, S. e.Organic synthesis in a modular robotic system driven by a chemical programming language.[2019]

35.  Faure, A. e.  Process control and scale-up of pharmaceutical wet granulation processes: a review. Eur. J. Pharm. Biopharm., [2001] 269-277.

36.  Landin, M.  Artificial intelligence tools for scaling up of high shear wet granulation process. J. Pharm. Sci. , [2017] 273–277.

37.  Das, M. a. ANN in pharmaceutical product and process development. In Artificial Neural Network for Drug Design, Delivery and Disposition. Elsevier, [2016] 277–293.

38.  kraft, D. (n.d.). System and methods for the production of personalized drug products.

39.  Gams, M. e. Integrating artificial and human intelligence into tablet. AAPS PharmSciTech , [2014] 1447–1453.

40.  Aksu, B. e. A quality by design approach using artificial intelligence techniques to control the critical quality attributes of ramipril tablets manufactured by wet granulation. Pharm. Dev. Technol. , [2013] 236–245.

41.  Dragoi, E. a. On The use of artificial neutral networks to monitor a pharmaceutical freez-drying process. Drying Technology, [2013]72-81.

42.  Available online. (2022, December 6)

43.  Mag.k.k, P. M. Artifial Inteligence in drug development: Present Status and future Prospects. Drug discovery Today, [2019]773-80.