A Landscape of Artificial Intelligence (AI) In Pharmaceutical R&D
This market research report aims at providing a “bird’s view” on the emerging ecosystem of AI-based technology companies (primarily, startups) focused on drug discovery and development. All the graphs and diagrams in this report below are dynamically generated and updated based on a manually curated database of companies. This report is based solely on the publicly available data from Internet. Please, send any suggestions and comments about its contents to email@example.com.
Customize report: uncheck boxes to hide redundant graphs
The below diagram shows how many of AI-based drug discovery startups emerge each year, and how much VC money they get collectively (only publicly announced rounds included).
Another important indicator of pharma industry’s willingness to adopt AI innovations is the number of collaborations between prominent pharma players and AI-technology vendors (startups). As one can see, the number of collaborations has been growing so far to reach the maximum in 2018.
(Note, you can click on any datapoint to see the list of AI-startups emerged in that year.)
This connectivity chart below shows some of the key industry collaborations between pharma companies and AI-driven technology vendors. It should be noted that technology giants, such as Google, or IBM are actively entering drug discovery/pharma R&D space either by direct partnerships, or via investing VC money into drug discovery startups.
The below chart shows how the AI-driven startups included in this study are grouped by their relative expertise in chemistry vs. biology. Understandably, most drug discovery research organisations take into account both aspect, but here we discuss the core competency.
The AI-driven drug discovery ecosystem is primarily focused in the US according to the diagram below. However, this perceived geographical disproportion is largely due to the fact fact, that many companies, for example, Chinese startups, are actually incorporated in the US. So here we assign geography formally, by the place of incorporation.
The next two diagrams provide a general idea of the research approach that is typical for AI-driven drug discovery startups and technology vendors.
Interestingly, most AI-innovations are focused around small molecules as starting points for drug discovery, employing target-based drug discovery (TBDD) paradigm. This is, probably, predictable as ligand-target interactions are easier to describe and fit into a typical machine learning/AI-driven modeling process. compared to other approaches, such as phenotypic screening.
AI startups are distributed across various therapeutic areas in a more or less the same way as any other emerging biotech startups — oncology is a major dominant as the most attractive area for drug discovery from many points of view.
To review the catalog of AI-technology vendors in the context of various drug discovery stages, please check “The AI Map of Drug Discovery”.
Next, it is seen that AI-technology vendors are involved in almost every research aspect of modern drug discovery and development process — from data mining and biology research all the way to helping organize, manage, and improve clinical trials. While some technology vendors focus on a narrow area, like de-novo design or ADME predictions, others develop an end-to-end AI-driven platforms to deliver “ready-to-trial” drug candidates.
Finally, the companies were clustered conditionally by the primary type of data they use for training AI-models (highly overlapping in many cases).