A new medicine’s research and development process is complex and lengthy and requires significant investments. The pharmaceutical industry finds it increasingly hard to develop and market efficient medication. The entire process can last up to 12 years and cost between 1.9 and 3.2 billion dollars, according to DiMasi et al., Journal of Health Economics, January 2016.
Doctors carry out complex biochemical and physicochemical procedures during clinical trials using massive data. The probability of a drug passing all tests is less than 12%, according to CSIC sources.
The pharmaceutical industry has dived into computing, artificial intelligence (AI), and machine learning (ML) to improve this process.
AI and ML in the pharma industry
The biopharmaceutical industry has been one of the key beneficiaries of artificial intelligence, mainly because it helps streamline the research and development of medicine and reduce research costs and the number of drugs undergoing clinical trials.
The pharmaceutical industry usually collaborates with hospitals and tech companies. That was the case during the COVID-19 pandemic, where the sector used AI algorithms to search for information, conduct studies, and carry out test simulations faster.
These algorithms helped identify the gene coding for target proteins that could bind favorably to drugs, reducing the initial drug development phase from around five years to only a few months.
These AI-powered techniques have played a vital role in developing COVID-19 vaccines with companies such as AstraZeneca or Janssen. Likewise, AI has made it easier to predict how the pandemic might continue to spread and which mistakes we can avoid if we encounter a similar situation.
The use of Big Data and AI in the healthcare industry is quite positive, impacting both pharma companies and society as a whole. That’s why many organizations have begun to bet on this technology.
The use of AI in drug research
Artificial intelligence in the biopharmaceutical industry has revolutionized the sector by facilitating access to biometric data agilely and helping identify hidden patterns.
AI may also prove helpful in these cases:
Proper patient segmentation is one of the biggest challenges in clinical trials. Choosing patients according to eligibility, suitability, motivation, and empowerment is crucial, and these processes usually involve significant delays that AI can solve.
Thanks to these mechanisms, a more efficient patient classification is possible, and appropriate treatments can be promptly administered, anticipating potential risks and getting more favorable results.
It’s possible to automate medical record screening or predict behavior patterns to understand how patients react to specific treatments or identify the reasons behind their hospital readmission.
With artificial intelligence, hospitals and healthcare institutions can reduce the number of in-person visits and replace them with online appointments. AI also allows healthcare professionals to monitor patients, regardless of whether they visit the healthcare center in person.
Radiologists can also benefit from these innovations. Artificial intelligence facilitates faster image diagnoses, streamlining the initial stage of treatments.
Pharmaceutical compound research
AI can automate complex tasks and save time.
During the initial stages of medicine research, AI significantly facilitates cellular trial analysis, molecular structure modeling, as well as the prediction of physicochemical properties of compounds, among other tasks.
Artificial intelligence also helps understand the possible interaction between proteins and ligands, allowing for creating molecules with more significant potential.
It’s also helpful when it comes to identifying and differentiating images, probably two of the least efficient processes in the development of medicine. Learning machines can spot tiny differences in cellular structures from microscopic images.
AI not only streamlines the initial stages of the process. It also proves valuable for quality control and optimizing medication administration.
Drug reuse or repositioning
Drug reuse is a strategy that seeks to discover new uses for medication that has already been approved.
That’s the case with Botox, for instance. It was initially created as a treatment for strabismus, but it was later discovered that it could also help with migraines and remove wrinkles.
Reusing drugs involves fewer risks and streamlines the development process. However, effectively combining clinical trials can be expensive and time-consuming.
In this context, artificial intelligence can develop hypotheses faster, accelerating clinical trials.
Advantages of implementing artificial intelligence
Artificial intelligence benefits significantly by predicting how molecules might react to certain medications.
It reduces the failure rate of clinical trials
Drugs in clinical trials have a success rate of under 12%. The reasons behind that figure include the lack of efficiency, adverse effects on humans, and the high costs of resources, all of which hampers the commercialization of the product.
Thanks to advanced artificial intelligence technology, the success rate of molecules can be increased, thus reducing the medicine’s adverse effects on humans. Machine learning algorithms can also improve efficiency without compromising security.
It accelerates the research of new medicines while reducing costs
The drug development process can last up to 12 years, including the preclinical and clinical stages, and it costs more than 2 billion dollars.
As new issues arise in different stages of the development process, the price continues to go up, especially during phase 3, where biomarkers differentiate drug responses.
Only the search for the compound —which involves target and lead selection, validation, and optimization— can last between 4 and 6 years. And at this point, only 1% of compounds reach the following phase.
That’s why using AI and ML in the pharma industry is crucial to reduce the risk of failure, particularly at the beginning of the process. A streamlined research and development process helps save time and resources.
It helps identify innovative treatments
Thanks to machine learning, technological advances and innovation in the pharmaceutical field have skyrocketed.
One of the critical developments powered by artificial intelligence involves personalizing treatments to adjust to each patient’s specific needs.
How companies are using AI and ML in the pharma industry
Many companies are already implementing artificial intelligence to save time and money in drug development.
Although research professionals are still vital in these processes, machine learning can significantly benefit pharma companies.
Moderna is an excellent example of artificial intelligence implementation. During the COVID-19 crisis, many wondered how the company manufactured the vaccine much faster than usual.
This achievement was due to Moderna’s use of state-of-the-art software and algorithms.
Exscientia and Sumitomo Dainippon Pharma
This case is one of the most striking ones, as it’s the first AI-created drug tested by humans. This is the case of molecule DSP-1181, which doctors developed with the help of AI as a long-acting serotonin 5-HT1A receptor antagonist, a drug used to treat an obsessive-compulsive disorder.
The development of this molecule was possible thanks to automated learning processes that allowed algorithms to generate millions of molecules and filter them to get the right one.
This Canadian company used artificial intelligence to identify the genetic cause of some diseases. Thanks to AI, developing drugs that regulate the defective genes causing these diseases is possible.
Many pharmaceutical companies have joined the digitalization era and are betting on artificial intelligence. While it’s still too early to know its future impact on drugs, the current results are satisfactory.
Our Data & AI Studio and Healthcare and Life Sciences Studio aim to bridge the gap to help pharma organizations achieve their mission of delivering innovation and services faster and more efficiently.