Expanding upon the key observations in our previous reports with new knowledge and analytics of Q2 2019, we can now better distinguish main trends within the industry in Q3 that will be shaping the market of AI in Drug Discovery in 2020 and beyond. 

 

Following a long lasting period of skepsis in 2018 and previous years the industry continues the “heating up” trend observed in Q2 also in Q3. This trend mainly consists of substantial increase in the volume of investments (notably, venture rounds become larger), financial support and the number of collaborations in Q3 of 2020. The industry’s growth dynamics is mainly influenced by the more active participation of largest pharmaceutical corporations in the AI-related investment and research collaborations. The number of research collaborations between pharma companies and AI-expertise vendors continues to increase. 

 

Despite the fact that IT and Tech corporations are more advanced in the AI tech in comparison to Pharma Corporations, they are increasingly open to co-operations and other forms of partnership with conventional drug makers -- in order to leverage their computational infrastructures and high-tech opportunities with enormous experience of pharma/biotech companies, their gigantic collections of datasets and novel approaches. AI here helps to discover new drugs, molecules, repurpose already existing drugs, find new targets that influence the illness progression etc. In this case, AI-Companies entering healthcare space are seeking strategic benefits, such as the ability to co-own important scientific discoveries and intellectual property obtained from the partnership and ability to continue developing new algorithms that will help these companies to leave their mark on medical history.

 

As 2020 marks a challenge in the ability to innovate, transform and adopt AI at scale faster, the so-called “Big Gap” continues narrowing due to rapidly increasing attention and activity of pharmaceutical companies with regards to AI prospects. Even tech giants like Google and Tencent are willing to expand their super-platforms to the area of pharmaceutical research. Having much bigger expertise of building and integrating super-platforms, currently they are conducting significant M&As and gaining some expertise in the area of the drug discovery, which would enable in the nearest future their expansion in this area. At the same time, the number of the deals between BioPharma corporations and AI companies aiming at the application of AI in drug discovery increased comparing with the same period in 2019. 

​

Global shortage of AI talent continues to be a serious  challenge for Biopharma industry, repeating the trend from our previous reports. However it should be noted that there is an increase in number of training courses and overall representing of AI-related directions in education programmes worldwide. Big pharmaceutical companies invest huge amounts of money in preparing of such specialists. But still the majority of talented AI professionals have been acquired by traditional IT-corporations and have been applied for purposes other than AI in healthcare. Therefore a lack of experienced specialists to support the activities of AI for Drug Discovery companies in particular is still a matter of today’s reality. Consequently large pharmaceutical companies continuously increase competing for the talented AI specialists as a valuable resource. Even specialized AI-driven drug discovery companies cannot fulfill gaps of AI talents as only 15.6% of their stuff being AI-experts.

 

Technologies, based on Deep learning (DL) algorithms will hold their leading position in the pharmaceutical AI race. Generative Adversarial Networks (GANs) and their variants are being increasingly regarded as a “golden standard” of innovation in the pharmaceutical AI space. 

 

Lack of available quality data is still a challenge for Investigations in AI and cooperating between AI and non-AI companies. The significant bottleneck in the AI applications for drug discovery purposes is the need to have correctly prepared, systematized and properly linked data that is ready to processing or is at least easy to manipulate with. Such types of data are quite scarce for the life sciences industry. A lot of research data in drug discovery is poorly validated and provided under a strict code of secrecy due to the high level of competition between drug makers. This is an issue unless there exists a well trained AI that is able to operate with unsorted collected information. THis means that as AI technologies evolve the weight of problems with unsorted information will decrease. 

 

There is a growing “AI democratization” trend, making machine learning and other advanced data analytics and modelling technologies increasingly commoditized, and available for use by non-AI experts. Examples include cloud-based “drag-and-drop” model builders, AI-focused frameworks, specialized out-of-the box software packages, and pretrained ML models. 

 

COVID19 accelerated progress in the pharmaceutical AI space, mainly due to the urgent need of running drug repurposing programs to come up with quick solutions to the pressing need, and the need to analyze large amounts of dynamically changing healthcare information (e.g. In the area of epidemiology, diagnostics and so on). The urgency of the situation stimulated accelerated research in the AI space and increased investments into such programs and projects.

 

Valuation of the industry is thought to show substantial growth, that, however, can be delayed in time. This appears to be a result of the general growth in the number of active business players, rather than an increase in the new products’ value.. No AI-derived drug has been approved by the FDA or validated in clinical trials so far. Despite this fact first milestones are expected to be reached by the end of 2020. On the other hand, the anticipated global financial crisis may hinder the industry’s growth dynamics, delay the AI adoption at scale, as well as the emergence of the first AI-derived blockbuster drugs.  

 

AI-friendly decision makers in pharmaceutical companies will become a great advantage in the competition for faster and cheaper development of new drugs. According to a recently conducted research, US, Japan and Germany are the countries with the biggest concentration of AI-friendly decision makers. About 46% of them work in pharmaceutical industry, while  only 3% are dealing with both AI and Pharma. In such more efficient application of AI technologies are expected to be observed and it will bring faster and more noticeable results.