Small molecules set to dominate pipelines with the help of AI

By Ben Hargreaves

- Last updated on GMT

(Image: Getty/Rost-9D)
(Image: Getty/Rost-9D)

Related tags CAS AI Big data Drug discovery

Though larger pharma companies are investing heavily in biologics, the industry is set to be predominantly focused on small molecules moving forwards, claims CAS MD.

The rapid development of the cell and gene therapy space, which has seen approvals for chimeric antigen receptor (CAR)-T treatments​ and adeno-associated viral gene therapies​, has led to greater investment in the area.

As a result, individuals within the industry have called on the larger companies to still invest into the small molecule space​ rather than place too much emphasis on biopharmaceuticals.

The call for broader investment comes as the growth of drug development is beginning to move towards large molecules​.

Todd Wills, business leader at CAS

However, when in-PharmaTechnologist (IPT​) spoke to Todd Wills (TW​), business leader of CAS, he explained that trends observed by his company suggest that there is still a large focus on small molecules within the discovery pipeline – further than this, the introduction of AI and big data technology could emphasise this.

IPT: Could you outline the work CAS does?

TW:​ CAS initially started more than 110 years ago as an information aggregator, collecting chemistry abstracts and data from journals. Over the course of our history, we've grown along with the publishing market and continued to aggregate as much chemistry-related content as we can. We've essentially created, and curated, a database of chemistry content for industry, including the pharma and biotech industries.

IPT: What services are provided to those industries?

TW:​ We're helping drug hunters chart a course into the chemical space. In order to do that, we started by mapping what is already known, from a chemistry perspective, and that's where we came up with an idea for measuring innovation based on the number of people who have explored certain regions of chemical space. By knowing what has already been explored, we can discover what areas of chemistry are open for further examination; we can help drug hunters decide where they want to go and we can help them get there, with some of our retro-synthesis offerings.

IPT: How does this work in practice?

TW:​ Companies tell us about a compound of interest and we'll give them pointers to other compounds that are available to be patented and developed into a drug. Or, they can tell us a target they're interested in and we'll go out into the chemical space then point them to pockets of compounds that could be biologically active against a particular target. That's where the big data comes in, we use the known chemical space to identify descriptors at a molecular level to then predict what molecules, which are not patented, might be the most promising for them to develop a drug around. 

IPT: What kind of trends have you noticed, in terms of recent drug approvals?

TW:​ There are some interesting insights on recent FDA approvals. If we look at the 2018 FDA drug approval list, there's a couple of interesting drugs that pop out because they show up in a chemical space that has not been very highly explored. There are also drugs approved that show up in the chemical space that are highly explored.

An example is Tibsovo (ivosidenib), which is in an area not very well explored. If you look at the number of other compounds on that same framework, there's six other known compounds – all were disclosed by Agios Pharmaceuticals patents. It's an example of a company going into an unknown space and staking claim to a specific region.

IPT: What type of compounds are the focus of drug pipelines?

TW:​ At a high level, small molecules continue to be the major source of new drugs and if you look at the last five or ten years, you can see that trend. If you compare that to the current late-stage trial pipeline, it looks like that relationship continues to hold – based on the number of drugs going through Phase III trials.

IPT: Will the use of AI discovery change this?

TW:​ There's been a lot of articles covering the amount of venture capital into AI-related companies, which are focused primarily on the small molecule end of things. I think this is because small molecules are easier to apply big data to than biologics – people are starting with the low hanging fruit.

IPT: What has been the reaction of companies working with you to explore new regions?

TW:​ I think it's opened a lot of people's eyes, from two different angles: one is that they didn't realise we could do this and the other aspect is some of the things we're proposing is a slight twist on current methods of drug discovery. They know that they have biases in the way they operate and they see this method as a 'sanity check'.

Todd Wills is business leader for CAS, with a responsibility for the technology-enabled consulting business – serving clients in the Pharma, Biotech, Agro, and Diversified Chemical markets. Prior to this role, he worked as a business leader at Emerson Network Power.

Related topics Markets & Regulations

Related news

Show more