AI and Pharma Industry: A New Love Story

AI and Pharma Industry: A New Love Story

The medical industry has become a Research and Development (R&D) sector where Artificial Intelligence (AI) has made its entry in a very powerful manner. They are many signs that we are moving more and more toward speciality AI applications. The quality of any AI applications is directly related to the quality and consistency of the data that they are fed with.

One of the most important areas of progress in AI will be the quality and consistency of available data. Therefore, the Internet of Things (IoT), and its ability to capture orderly and standardised data, will be the next revolution.

To this date the major medical areas affected by AI have generally been:

·       Robotic Surgery: For example, laser eye surgery and hair transplants are typical procedures that are simple enough for effective treatments.

•        Image Analysis: AI automated systems are assisting experts in examining X-rays, retina scans and other images.
•        Genetic Analysis: As genome scans have become a routine part of medicine, AI tools that quickly draw insights from the data are essential tools.
•        Pathology: Experimental systems have so far proved adept at analysing biopsy samples. Next step will be to get them approved for clinical use.
•        Clinical Decision Support: A new ground that has, somehow, not yet proven its value. A typical example of this technology will be for predicting septic shock.
•        Virtual Nursing: Systems can check on patients between office visits and provide automatic alerts to doctors.
•        Medical Administration: New AI-enabled tools can increase efficiency in tasks like billing and insurance claims.
•        Mental Health: Mining mobile phone and social media data can be used by researches for monitoring depression.

In the last few years, we have seen many IT giants (IBM, Apple, Google, Amazon, etc…) working on AI solutions for the medical industries. Not all of them have, however, been successful so far for different reasons. At a 2017 conference of health IT professionals, IBM CEO Ginni Rometty told the crowd that AI “is real, it’s mainstream, it’s here, and it can change almost everything about health care,” and added that it could usher in a medical “golden age.” Experts in computer sciences agree with her as indeed AI has the potential to transform the health care industry. Ironically, although possibilities have been demonstrated in carefully controlled experiments, so far only a few AI-based tools have been approved by regulators for use in hospitals and doctors’ offices.

IBM AI multi-purpose solution, called Watson (an IBM supercomputer that combines AI and sophisticated analytical software for optimal performance as a “question answering” machine), failed to deploy significant applications that would have made a great contribution to the medical industry. Eliot Siegel, a professor of radiology and vice chair of information systems at the University of Maryland, collaborated with IBM on the diagnostic research. While he thinks AI-enabled tools will be indispensable to doctors within a decade, he eventually stated that he was not confident that IBM will build them. “I don’t think they’re on the cutting edge of AI,” says Siegel. “The most exciting things are going on at Google, Apple, and Amazon.”

As for Martin Kohn, who originally came to IBM with a medical degree from Harvard University and an engineering degree from MIT, was excited to help Watson tackle the language of medicine, thus creating some sort of superdoctor, left IBM in 2014 saying that the company fell into a common trap: “Merely proving that you have powerful technology is not sufficient,” he says. “Prove to me that it will actually do something useful—that it will make my life better, and my patients’ lives better.”

It is a well now fact in the medical industry that the search for new drugs is a long and costly process. Even with the use of AI, innovation can be a very difficult path to follow. However, a French biotech company, Pharnext, established in 2007 has just achieved a major milestone in using AI for creating new treatments and has become a model of a successful AI biotech company. The success of this company is partly due to the fact that Pharnext uses quality, not raw, data that is reliable and proven. Their database is complete enough as it needs for new data is low. Pharnext ingenious technique is also to work on molecules already commercialised to build molecular combinations. The benefits of this approach are many:

• The amount of data is controlled
• The quality of the molecules being screened is already established

·       As their databases are based on old molecules, the information is often of good quality and has been verified, allowing for high quality and good data formatting. Every day, molecules fall into the public domain. Pharnext approach can therefore restart tests with new molecules regularly. As time goes by, this process will be greatly mastered and perfected on an on-going basis. Potentially, Pharnext can do silico research and discover new treatment options.

·       It allows a medical project to reduce time from inception to market delivery by 5 years when compared to traditional research methods. This mean that a 15 years project is cut down to 10 years, thus achieving a 33% reduction.

Pharnext, therefore, has cracked the AI medical challenge by solely using what can be perceived as old data. This approach and the avoidance of new unvalidated data had led Pharnext to create a robust data environment for research purposes. The use of complex computer modelling helped Pharnext to naturally turn to in-silico research (analysis and experimentations carried out in a computer environment), rather than in the laboratory, that greatly accelerate the preclinical phases.  As they explain on their website: from all biological data associated with a given disease, Pharnext builds the molecular network of this disease, which represents an inventory of potential therapeutic target. From this network, Pharnext selects and repositions therapeutic molecules known for different indications, at new, lower doses, to identify New combinations Synergistic. Thus, active ingredients, which have worked in other diseases, find new use in other cases. A single protein has several functions. Pharnext gave a name to this technique: pleotherapy. It even filed it and protected it for its commercial exploitation. Even if pleotherapy is a novelty, the base is solid and proven. Researchers, then, do not work blindly and they are not reinventing new drugs, they are recombining and improving existing ones! By combining several molecules, they seek to find new therapeutic cocktails that offer effective synergies within the framework of specific diseases. Pharnext understood very well that in this little game of combinations, the AI was a champion. Unlike classical pharmacology, which focuses on the most accurate target possible, pleotherapy targets several mechanisms simultaneously.

Pharnext success allowed them to create partnership with other biotech companies. Galapagos, a Belgian pharmaceutical company, worked with Pharnext to explore improvements, create a new pipeline combination of synergistic drugs covering a wide range of indications in the inflammatory and neurodegenerative diseases. "These new combinations are centred around Galapagos' candidate drugs in the inflammatory field, which will be associated, thanks to our patented technology, with other molecules already marketed without a patent," Daniel Cohen told Les Echos in 2017.

Furthermore, Pharnext also partnered with Tasly in China. Tasly is a giant in the field of traditional medicines. Tasly employs more than 10,000 people and is part of the of the 10 largest companies in the sector. Here, the partnership is twofold:

•        Tasly will open the Chinese market to Pharnext drugs
•        Pharnext puts at Tasly's service technology as part of a common research platform

Yan Kaijing, CEO of Tasly Pharmaceuticals, stated: "Building on both the benefits of Tasly's research and development platform and its dense implementation in the Chinese hospital network; and on the remarkable technology Pharnext's drug discovery, we will develop combinations of Pharnext high-potential drugs to meet unmet medical needs. Supported pharmacology of biological disease networks. This partnership will allow us to harness the immense potential of medicines from traditional Chinese medicine to modernised medicine, thanks to a precise and new characterisation of the mechanism of action of each combination”.

The future of AI in the biotech and medical industry is thrilling as it is now achieving a major technological transition from what is fundamentally a difficult, administrative process ridden and unpredictable research industry to a more matured one that will benefit doctors, medical and paramedical support agencies, hospitals, drug research companies and ultimately patients.