AI in Pharmacy: Saving Billions of Dollars

Although Artificial Intelligence (AI) in pharmacy industry is still in its infancy, the process has already started and predictions for the near future are very optimistic. Definitely, the juice is worth the squeeze – AI can make the drug discovery process quicker, cheaper and more effective. AI really does have the necessary potential to change the entire process of drug development. At stake are billions of dollars of savings and… human health.

Currently, the estimated cost of single drug development, from discovery, through clinical tests to finally putting it on sale is at least 350 million USD. And it takes around 10 years to introduce the new medicine to the market. It is said that just 1 out of 10 drugs will successfully finish that long and winding process. So no wonder that pharmaceutical companies are actively searching for ways to shorten, cheapen and make the entire process more effective. And this is the place where modern technology – machine writing, analytics, and data science solutions come into play.

This is not science-fiction nor distant future. Technology is at hand! For example, in oncology, using AI algorithms combined with an individual patient’s data and his or her medical history, the computer can review the available treatment alternatives and recommend the most appropriate medicine combination.

Experts see the great potential for AI in pharmacy, especially in the two main areas:

•drug design

•predicting treatment results

Let’s focus now on each of these areas and find out how AI in pharmacy can be beneficial for the

business.

Drug design with AI analytics

The first step in drug development is to understand the biological origin of a given disease and its resistance mechanisms. However, in the traditional process, it is very challenging to integrate a large amount of data sources — and then find the relevant patterns. Machine learning algorithms can do that easily and analyze thousands of sources just in seconds. And that happens already! MIT scientists have created a system that analyzes available science data and on that basis makes a compound. Researchers have tested the device and, with it, managed to make 15 medicines. Almost without human assistance! For example, they have managed to create a machine-made aspirin, and four non-steroidal anti-inflammatory drugs.

Another challenge the scientists are facing every time they develop a new medicine is clinical tests. In order to succeed you have to find suitable candidates – first animals and then humans. AI with its machine learning can considerably speed up the process of looking for these candidates and rejecting the wrong ones. AI analytics can analyze genetic information to identify the proper patients for the tests. AI/ML can also reduce data errors, such as duplicate entries. AI algorithms, therefore, can indicate the perfect candidates for the clinical tests. As a result, it reduces animal testing and human clinical trials in pharmacy, because more data can be gained without their participation.

AI algorithms in pharmacy

What’s more, AI is very helpful in data analytics. Machine learning algorithms can screen hundreds of thousands of molecules in the human body to find biomarkers – particles that with 100% certainty indicate if you have given disease. That makes the process quicker and cheaper.

One of the examples of the AI algorithm in the pharmacy industry is the AtomNet. A deep learning neural network application made to help in drug discovery. The company behind the project says it is based on CNN technology (Convolutional Neural Networks) and helps by combining atoms to come up with possible molecules.

Here is the place for a small digression: nature is lazy. Atoms are, in general, not willing to combine with one another. Sometimes that happens easily and sometimes it is very difficult to combine two or more different atoms in one molecule, for instance in some cases you have to create a magnetic field to make it possible. So it takes time to find appropriate combinations of atoms. A lot of time. The company says that its algorithms are “trained” to analyze billions of atoms to find which will bond together to finally create an efficient compound.

The strategy of creating desirable compounds is called Release (Reinforcement Learning for Structural Evolution) and fulfills two tasks. First, it helps in the creation of compounds optimized for a specific characteristic. And second, it reduces the unwanted side effects of a medicine.

To sum up – machine learning and AI in pharmacy is undisputedly a great milestone. In 10-20 years from now, the process of creating a medicine will be faster, cheaper and more accurate. Predictions are that AI will shorten the time to create a drug just to several weeks instead of years.

Predicting treatment results with AI

Predicting treatment result is a second field in which AI will be helpful in pharmacy. Even now there are available data science solutions that allow predictions as to whether given medicine or treatment will produce desired results. It’s called Reverse Engineering & Forward Simulation (REFS) and is also based on machine learning.

One of the companies implementing that solution is GNS Healthcare. They say, that REFS machine learning algorithms can “predict” given patient’s response to drug treatment. REFS analyzes data consisting of many factors such as the body’s ability to absorb the compound and metabolism. REFS algorithms are looking for the elements that might have an influence on a patient’s response to a given drug. Next, they start multiple simulations that, as a result, provide the best drug treatment. On that basis, a doctor can indicate the proper treatment and prescribe the optimal drugs.

AI is the future of the pharmacy

Implementation of AI in the pharmacy industry is by all means beneficial. It reduces the time needed for a drug to get approval by the drugs department and reach the market. It offers much more accurate treatment, considering each patient’s situation and circumstances. And finally, it minimalizes the risk of wrong medical diagnosis, because treatment is based on data from millions of patients worldwide.

It is obvious now – if you run a pharmaceutical company you have to get acquainted with things like machine learning, data analytics, and AI consulting. It’s not an option. It’s a necessity. Companies that have already started implementing AI in drug development will advance quickly. It is expected that AI will take over almost the entire Ramp in pharmacy within the next 20 years.

‍