- The Key Patents of 10 Blockbuster Drugs to Expire in the Next 5 Years
- New Data Shows Rising Colorectal Cancer Mortality in Young Adults
- Arginine-Driven Metabolic Reprogramming in Liver Cancer: A Potential Therapeutic Target
- CAR-T Cell Therapy Achieves 100% Sustained Remission in Lupus Patients
- First TIL Therapy for Malignant Melanoma Approved
- Japan 10th COVID Wave: COVID-19 Hospitalizations Remain High
Artificial intelligence will lead drug development in future?
- Japan: Over 10,000 Applications for Health Damage from COVID-19 Vaccines
- Antioxidant Supplementation May Accelerate Tumor Growth and Spread
- FDA has mandated a top-level black box warning for all marketed CAR-T therapies
- ‘Elixir of Immortality’ Nicotinamide Riboside (NR) Virtually No Effect
- Can people with high blood pressure eat peanuts?
- What is the difference between dopamine and dobutamine?
- How long can the patient live after heart stent surgery?
Artificial intelligence will lead drug development in future?
In the past few years, the digitization of data in the pharmaceutical industry has grown tremendously.
However, the challenge posed by digitization is how to apply these data to solve complex clinical problems. This motivates the use of artificial intelligence because it can process large amounts of data through enhanced automation.
Artificial intelligence is a technology-based system that includes various advanced tools and networks that can imitate human intelligence.
At the same time, it will not threaten to completely replace the existence of human beings.
Artificial intelligence uses systems and software that can interpret and learn input data to make independent decisions to achieve specific goals.
The application of artificial intelligence in the field of medicine is constantly expanding.
Artificial intelligence involves multiple method areas, such as reasoning, knowledge representation, and solution search, including the basic paradigm of machine learning ( ML ).
A subfield of ML is deep learning ( DL ), which involves artificial neural networks ( ANN ).
They consist of a set of interrelated and complex computational elements, involving the “perception” similar to human biological neurons, simulating the transmission of electrical impulses in the human brain.
Neural networks involve various types, including multilayer perceptron ( MLP ) networks, recurrent neural networks ( RNNs ), and convolutional neural networks ( CNNs ).
More complex forms include Kohonen network, RBF network, LVQ network, backpropagation network and ADALINE network.
The following figure summarizes an example of the method domain of artificial intelligence.
Artificial intelligence helps drug screening
The process of discovering and developing a chemical drug can take more than 10 years and cost an average of US$2.8 billion.
Even so, 90% of therapeutic molecules failed to pass phase II clinical trials and regulatory approvals.
Algorithms such as nearest neighbor algorithms, RF, extreme learning, SVMs and deep neural networks ( DNNs ) can be used for virtual screening ( VS ) based on synthetic feasibility , and can also predict activity and toxicity in the body.
Some large biopharmaceutical companies, such as Bayer, Roche, and Pfizer, have cooperated with IT companies to develop artificial intelligence platforms for discovering treatments in the fields of tumor immunology and cardiovascular diseases.
Prediction of physical and chemical properties
The physicochemical properties of drugs, such as solubility, partition coefficient ( logP ), ionization, and intrinsic permeability, will indirectly affect the pharmacokinetic properties and target receptors of the drug.
Therefore, it must be considered when designing new drugs. Different artificial intelligence tools can be used to predict physical and chemical properties.
For example, ML uses the large data set previously generated in the compound optimization process to train the program.
The algorithm of drug design includes molecular description, potential energy measurement, electron density around the molecule, and three-dimensional atomic coordinates.
A feasible molecule is generated through DNN to predict its properties.
Biological activity prediction
The efficacy of drug molecules depends on their affinity for the target protein or receptor. Drug molecules that do not have any interaction or affinity for the target protein will not provide a therapeutic response.
In some cases, the developed drug molecules may interact with unexpected proteins or receptors, causing toxicity. Therefore, drug targeted binding affinity ( DTBA ) is the key to predicting drug-target interaction.
Artificial intelligence-based methods can measure the binding affinity of drugs by considering the characteristics or similarities of drugs and their targets.
The feature-based interaction identifies the chemical composition of the drug and the target to determine the feature vector.
In contrast, similarity-based interactions consider the similarity between the drug and the target, and assume that similar drugs will interact with the same target.
Web applications such as ChemMapper and similar integration methods ( SEA ) can be used to predict drug-target interactions.
Many strategies involving ML and DL have been used to determine DTBA, such as KronRLS, SimBoost, DeepDTA and PADME. ML-based methods, such as KronRLS, evaluate the similarity between drugs and protein molecules to determine DTBA.
Similarly, SimBoost uses regression trees to predict DTBA, while considering feature-based and similarity-based interactions.
Predicting the toxicity of drug molecules is essential to avoid toxic effects.
Cell-based in vitro tests are often used as preliminary studies, followed by animal studies to determine the toxicity of compounds, increasing the cost of drug discovery.
Some web-based tools, such as LimTox, pkCSM, admetSAR, and Toxtree, can help reduce costs. Advanced artificial intelligence-based methods look for similarities between compounds or predict the toxicity of compounds based on input characteristics.
The Tox21 Data Challenge organized by the National Institutes of Health, the Environmental Protection Agency ( EPA ) and the US Food and Drug Administration ( FDA ) is an initiative to evaluate several computational techniques for predicting the toxicity of 12707 environmental compounds and drugs .
The ML algorithm named DeepTox stands out. It recognizes the static and dynamic characteristics in the chemical description of the molecule, such as molecular weight ( MW ) and van der Waals force, and can effectively predict the toxicity of the molecule based on the characteristics of the predefined 2500 toxic groups.
The different artificial intelligence tools used in drug discovery are shown in the table below.
Artificial intelligence helps drug design
Target protein structure prediction
In the process of developing chemical drugs, predicting the structure of the target protein is crucial for designing drug molecules.
Artificial intelligence can help structure-based drug discovery by predicting 3D protein structure, because the design must conform to the chemical environment of the target protein site, thereby helping to predict the compound’s impact on the target and safety considerations before synthesis or production.
AlphaFold, an artificial intelligence tool based on DNNs, analyzed the distance between adjacent amino acids and the corresponding angle of peptide bonds, predicted the three-dimensional structure of the target protein, and correctly predicted 25 out of 43 structures.
Drug-protein interaction prediction
The interaction between drugs and proteins plays a vital role in the success of treatment.
Predicting the interaction of drugs with receptors or proteins is critical to understanding the efficacy and effectiveness of drugs, allowing the reuse of drugs, and preventing polypharmacology
. Various artificial intelligence methods are very useful in accurately predicting ligand-protein interactions, ensuring better therapeutic effects.
The ability of artificial intelligence to predict drug-target interactions is also used to help change the use of existing drugs and avoid polypharmacology.
Changing the use of existing drugs can be directly used in the second phase of clinical trials. This also reduces expenses, as the cost of restarting existing drugs is approximately US$8.4 million compared to newly developed drug entities ( US$41.3 million ).
The “guilt association” method can be used to predict the innovative association between drugs and diseases, which is a knowledge-based or computationally driven network.
In computationally driven networks, the ML method is widely used, which utilizes technologies such as support vectors, neural networks, logistic regression, and DL.
Drug-protein interactions can also predict multi-pharmacological opportunities, which is the tendency of drug molecules to interact with multiple receptors, resulting in non-targeted adverse reactions.
Artificial intelligence can design a new molecule based on the basic principles of polypharmacology and help produce safer drug molecules.
Artificial intelligence platforms like SOM, coupled with the existing huge database, can be used to connect several compounds with numerous targets and non-targets.
Bayesian classifiers and SEA algorithms can be used to establish the relationship between the pharmacological characteristics of a drug and its possible targets.
De novo drug design
In the past few years, the method of designing drugs from scratch has been widely used in the design of drug molecules. The traditional ab initio design method is being replaced by the evolved DL method. The former has the disadvantages of complex synthetic routes and difficulty in predicting the biological activity of new molecules. Popova et al. developed reinforcement learning for structural evolution strategies for de novo drug synthesis, including generating and predicting DNNs to develop new compounds. Merk et al. also used generative AI models to design retinoic acid X and PPAR agonist molecules, which have ideal therapeutic effects without the need for complex rules. The author successfully designed five molecules, four of which showed good regulatory activity in cell detection. The de novo design of artificial intelligence molecules is beneficial to the pharmaceutical industry because it has various advantages, such as providing online learning and simultaneously optimizing the data that has been learned, as well as suggesting possible synthetic routes for compounds, so as to achieve rapid pilot design and development.
Artificial intelligence helps pharmaceutical product development
The discovery of a new drug molecule requires it to be subsequently combined in a suitable dosage form with the desired administration characteristics.
In this regard, artificial intelligence can replace the old trial and error method. With the help of QSPR, various calculation tools can solve the problems encountered in the field of formula design, such as stability problems, solubility, porosity, etc.
Decision support tools use a rule-based system to select the type, nature and quantity of excipients based on the physical and chemical properties of the drug, and operate through a feedback mechanism to monitor the entire process and modify it intermittently.
Artificial intelligence helps pharmaceutical manufacturing
With the increasing complexity of the manufacturing process, and the continuous improvement of the requirements for efficiency and better product quality, modern manufacturing systems are trying to impart human knowledge to machines and continuously change manufacturing practices.
The application of artificial intelligence in manufacturing can prove to be a boost to the pharmaceutical industry. Tools such as fluid dynamics calculation ( CFD ) use Reynolds average Navier-Stokes solver technology to study the effects of stirring and stress levels in different equipment ( such as stirred tanks ) to automate pharmaceutical operations.
Similar systems, such as direct numerical simulation and large eddy simulation, involve advanced methods for solving complex flow problems in pharmaceutical production.
Artificial intelligence helps quality control and quality assurance
The production of required products from raw materials includes the balance of various parameters.
Product quality control testing and maintenance of batch-to-batch consistency require manual intervention. In many cases, this may not be the best method, indicating the necessity of artificial intelligence at this stage.
The FDA revised the current Good Manufacturing Practices ( cGMP ) and introduced a “quality by design” method to understand the key operations and specific standards that control the final quality of drugs.
Artificial intelligence can also be used to monitor online manufacturing processes to meet the expected standards of products.
The freeze-drying process monitoring based on artificial neural network is adopted, and the adaptive evolution algorithm and the local search and back propagation algorithm are combined.
This can be used to predict the temperature and the thickness of the dried filter cake at a future point in time ( t+Δt ) under specific operating conditions , and ultimately help to check the quality of the final product.
In addition, data mining and various knowledge discovery technologies in the total quality management expert system can be used as valuable methods to make complex decisions and create new technologies for intelligent quality control.
Artificial intelligence helps clinical trial design
The purpose of clinical trials is to determine the safety and effectiveness of a drug under specific human disease conditions.
It takes 6-7 years and a lot of financial support. However, only one out of ten small molecules entering clinical trials may be successful, and the low success rate is a huge loss to the industry.
These failures may be caused by improper patient selection, insufficient technical requirements, and poor infrastructure.
However, there is a large amount of digital medical data available, and these faults can be reduced by the implementation of artificial intelligence.
The registration of patients requires one third of the clinical trial time.
The success of clinical trials can be guaranteed by recruiting suitable patients, otherwise it will lead to approximately 86% of failed cases.
AI can use patient-specific genome-exposure group analysis to help select specific disease populations for recruitment in the second and third phases of clinical trials, which helps early prediction of available drug targets for selected patients.
Preclinical discovery of molecules and the use of other aspects of artificial intelligence ( such as predictive ML and other reasoning techniques ) to predict lead compounds before the start of clinical trials can help early prediction of lead molecules that pass clinical trials and consider selected patient populations .
Patients who withdrew from clinical trials accounted for 30% of clinical trial failures, creating additional recruitment requirements for completing the trial, resulting in a waste of time and money.
This can be avoided by closely monitoring patients and helping them follow the expected protocol of the clinical trial.
The mobile software developed by AiCure monitored the routine drug intake of patients with schizophrenia in the second phase of the trial, which increased the compliance rate of patients by 25% and ensured the successful completion of clinical trials.
The market prospects of artificial intelligence in the pharmaceutical industry
In order to reduce the financial costs and failure probability associated with pharmaceutical development, pharmaceutical companies are turning to artificial intelligence.
The artificial intelligence market has increased from US$200 million in 2015 to US$700 million in 2018, and is expected to increase to US$5 billion by 2024.
From 2017 to 2024, it is expected to grow by 40%, which indicates that artificial intelligence may revolutionize the pharmaceutical and medical industries.
Many pharmaceutical companies have and are continuing to invest in artificial intelligence and are collaborating with artificial intelligence companies to develop necessary healthcare tools.
Incomplete statistics, foreign leading AI pharmaceutical platforms include: Schrödinger, Exscientia, AbCellera Biologics, Recursion Pharmaceuticals, Atomwise, Benevolent, Insilico, Silicon Health Pharmaceutical, Insitro, Cyclica, etc.
The ongoing challenge of adopting artificial intelligence
The overall success of artificial intelligence depends on the availability of large amounts of data, because these data are used to provide subsequent training for the system.
Accessing data from different database providers may incur additional costs for the company, and the data should also be reliable and high-quality to ensure accurate result predictions.
Other challenges that hinder the full adoption of artificial intelligence in the pharmaceutical industry include the lack of skilled personnel to operate artificial intelligence-based platforms, the limited budget of small organizations, fears of replacing humans leading to unemployment, skepticism of artificial intelligence-generated data, and black box phenomena.
Despite this, artificial intelligence has been adopted by many pharmaceutical companies, and it is estimated that by 2022, the pharmaceutical industry will generate $2.199 billion in revenue through artificial intelligence-based solutions.
Pharmaceutical organizations need to understand the potential of artificial intelligence technology in solving problems and understand the reasonable goals that can be achieved.
Only with skilled data scientists, software engineers who have a full understanding of artificial intelligence technology, and a clear understanding of the company’s business goals and R&D goals can the potential of the artificial intelligence platform be fully utilized.
The advancement of artificial intelligence is constantly working to reduce the challenges faced by pharmaceutical companies, affecting the drug development process and the entire product life cycle, which is reflected in the increase in the number of start-ups in the industry.
The current healthcare sector is facing some complex challenges, such as the increase in drug and treatment costs, and society needs to make specific and major changes in this area.
With the application of artificial intelligence in the manufacture of medical products, personalized medicines with required dosages, release parameters, and other required aspects can be manufactured according to the needs of individual patients.
Using the latest artificial intelligence-based technology can not only speed up the time required for product launches, but also improve product quality and overall safety of the production process, and make better use of existing resources while improving cost-effectiveness.
For the application of these technologies, the most worrying is the subsequent unemployment and the strict regulations required to implement artificial intelligence.
However, these systems are just to make work easier, not to completely replace humans.
Artificial intelligence not only helps to identify suitable compounds quickly and without obstacles, but also helps to provide suggestions for the synthetic routes of these molecules, as well as the prediction of the required chemical structure, as well as the understanding of drug-target interactions and their SAR .
Artificial intelligence can also make a major contribution to further incorporating the developed drug into its correct dosage form and optimizing it.
In addition, it can also help rapid decision-making, thereby speeding up the production of better quality products, while ensuring consistency between batches. Artificial intelligence can also help determine the safety and effectiveness of products in clinical trials, and ensure the proper positioning and cost of products in the market through comprehensive market analysis and forecasting.
Although there are no drugs developed using artificial intelligence-based methods on the market, and there are still some specific challenges in implementing this technology, artificial intelligence is likely to become an invaluable tool in the pharmaceutical industry in the near future.
1.Artificial intelligence in drug discovery and development. DrugDiscov Today. 2020 Oct 21;S1359-6446(20)30425-6.
Artificial intelligence will lead drug development in future?
(source:internet, reference only)
Important Note: The information provided is for informational purposes only and should not be considered as medical advice.