In Episode 66 of 17 Minutes of Science, we are joined by Dr. Marius Galyan of Galyan Bio to talk about how he is pioneering the next generation of drug discovery. Dr. Galyan is a trained neurologist with a PhD in molecular oncology from the University of Munich. In his academic career in Saint Gallen, Switzerland, he led CNS clinical trials in Multiple sclerosis with a main focus on cognitive impairment.
Dr Sarah Cheeseman: Well, hello, everyone, welcome to 17 Minutes of Science, my name is Sarah Cheeseman and I'm a technical solution scientist at InVivo Biosystems. Happy New Year! It's good to be here with you, even if things are kind of an unusual time as we begin 2022. But I am delighted to have my guest with me here today, Dr. Marius Galyan. Marius is the CEO for a company called Galyan Bio, his namesake, and one of the core missions of their company is pioneering the next generation of drug discovery for neurodegenerative diseases, which is obviously important and lofty goal. So Marius is a trained neurologist. He has a doctorate in molecular oncology from the University of Munich and in his academic career in St. Gallen, Switzerland. He led CNS clinical trials in multiple sclerosis with a main focus on cognitive impairment. So please welcome Marius, and I'll let you say a few words I'm going to start my timer and we're going to go for it.
Dr Marius Galyan: Thank you for that.
Dr Sarah Cheeseman: Well, it's good to have you here. So we'd love to start by hearing a bit more about you, about your company and how things are going.
Dr Marius Galyan: So yeah. I started Galyan Bio in 2015 with a research idea to discover new drugs for neurodegenerative diseases, and Galyan Bio's mission is to deliver breakthroughs in changing the future of aging and halting the progression of neurodegenerative diseases, especially Huntington's disease, Parkinson's and Alzheimer's. Where the medical need is huge. And the first compounds we developed were for Huntington's disease. And this disease is - we know a lot about the genetic mutations and causes of the disease. Thus, we have good animal models reflecting human pathology.
Dr Sarah Cheeseman: Mm-hmm. And so did you - did you set about to develop novel compounds?
Dr Marius Galyan: Yes, we set out to discover new compounds, new approaches, because for the last 30, 40 years, nobody was able to find disease modifying drugs for these indications. We have only symptomatic treatments and I thought this is unsatisfactory, especially as a clinician. I saw in my clinical practice patients where I could not help them to stop the disease, just to help them dealing with the symptoms. But I could not stop the disease developing, progressing forward.
Dr Sarah Cheeseman: Hmm. And why do you think in all your clinical experience, this group of diseases, I'm going to just broadly say neurodegenerative diseases are so difficult to treat and work with?
Dr Marius Galyan: Yes, I think it's mainly because they are heterogeneous diseases where there are several factors pathological pathways which are involved, for example, in Alzheimer's, we have, of course, that amyloid-beta but also pathological tau accumulating. But on top of it, we have vascular disease and also inflammation playing a role. So there are a lot of different pathological pathways involved. And if you address just one of these pathways, it does not - it is not sufficient. And therefore, we need treatments which address several pathways at once and which are very upstream, which work in the beginning, where the progress - where the disease progresses forward and destroys the neurons. And this is, I think, a better approach.
Dr Sarah Cheeseman: Hmm. Ok, that's a great segue into my next question, which is using animal models in this process of discovery. So I know you use more than one kind of animal model in your work to advance your studies. And one of them, of course, is C. elegans, which is how we to know one another because InVivo Biosystems works with that. Can you tell us a bit more about how you've chosen the models and how you've worked with them and how they maybe complement one another, what they each of them bring to the problem?
Dr Marius Galyan: Yes, C. elegans is really ideal to generate fast results. That was the primary reason why I selected that model. And we have the opportunity to validate cost efficiently our compounds. And another advantage is that C. elegans is widely used for longevity studies, and we have more experience interpreting correctly the results and comparing with other compounds in development for aging. And that was the main reasons why I selected that model. And in the next step, we will finish optimization of our longevity drug and test them in mouse aging models. So we are using the data which we generated at your company to develop our program further and use that data to direct us in the right direction.
Dr Sarah Cheeseman: Ok. So kind of a stepwise progression - lets you validate. Yes, we hear that from from a number of people that that's a useful way to progress quickly by using a simpler model and then taking it up to a mammalian system. Well, we'll stay tuned to to hear how the worm model has informed what you learn in the rodent models. I understand that that Galyan Bio also incorporates a lot of artificial intelligence into your drug discovery process. Can you tell us more about how you built neuro AI and what it does?
Dr Marius Galyan: Yes, it started all in 2017 with an EU grant. We set up a new in silico drug screening platform. In the following years we gradually improved our procedures, and we harnessed the advanced AI methods for the new drug design to the small molecule development and aid in drug safety profiling. And our in-house data sets are combined with our advanced AI platform and empower the development process to identify effective medicines for neurodegenerative diseases. And we screened in the beginning compounds in silico for binding on new protein surfaces and use neural networks to identify chemical fragments based on only target crystallography interaction data. So we use the known data and use that in our neural networks to identify new infection partners and neuro AI is optimized to address novel therapeutic targets, including difficult to address protein actions and Galyan Bio has developed machine learning algorithms to identify compounds that could potentially interact with secondary targets as well responsible for toxicity. So we are using this method to identify potential compounds which could have toxic issues, and so we select them and do not use them further, and thus we are reducing the costs. So, and, whilst we are eliminating compounds before synthesizing them and testing them in various time consuming assays. And the next step, we will further progress our AI platform, of course, we will integrate in our system computational sorting out and through, and compare various properties of the compounds as well. So this will be the next step to do that in silico and identify potential small molecules which will further streamline and speed up the process.
Dr Sarah Cheeseman: Hmm. So can you give us a sense of - just as you're speaking, I'm thinking of that is an enormous amount of data to go through. How many compounds do you put into that, for lack of a better word, that neuro AI platform for the sort of initial scan of feasibility?
Dr Marius Galyan: So we have interaction data for about 20,000 compounds, ligands with their targets and these ligands, we are fragmenting them. So our database is larger than that because we are taking these ligands and building fragments of them and with their interaction partners. So in the end, we have several hundred thousand data interaction fragments with their protein interaction partners and this database with what we use to generate new compounds for new targets, then.
Dr Sarah Cheeseman: Is it, in your experience and with your colleagues in the field, is this becoming the standard practice for developing new therapies - to do this as the first step?
Dr Marius Galyan: Yes, I think more and more companies are using that and in traditional drug discovery process, the initial drug candidate took four to five years. With AI platforms, if they are fully developed, it can only take several months. And that's a huge factor, and thus, the development costs can be reduced. And our goal is to lower failure rates and ultimately reduce also the price of drugs because currently the drug discovery process is arduous and unsustainable and the cost of drugs are too high. And if we want to develop more and more drugs, we we need a more cost efficient methods to reduce also the drug costs.
Dr Sarah Cheeseman: Yes. I'm also thinking, though, that with this methodology, you can just screen so many more potentials, so versus the more traditional process where you might miss some things that actually could be really useful.
Dr Marius Galyan: Yes, and that's one of the problems because we - and that's the approaches we will use in future, is there are hundreds of known targets for oncology. There are a lot of targets for which we don't have any drugs, so they are readily available validated drug targets for which no compounds are existing. So the list of potential targets is huge. And so we can use these new approaches to identify and do it quickly and get to clinical candidates and test them and see if these targets are really in the clinical field useful to improve patient outcomes.
Dr Sarah Cheeseman: Hmm. Fascinating. And so to bring it down to a specific example, one of the drugs you discovered for Huntington's diseases recently entered a clinical trial. Did that go through this process you just described? Did you take an A.I. approach for that compound?
Dr Marius Galyan: So our compound is not yet in the clinic, so it is a clinical candidate where we have to do still some safety, but it's close to the clinic. So and this drug has been developed with the A.I. system correctly, yeah.
Dr Sarah Cheeseman: Mm-hmm. And so what is the being the timeframe for that project?
Dr Marius Galyan: Yes, it is a little bit. The discovery of the compounds and getting to the first scaffold, which was the clinical candidate in in the end, was under year. But because of funding reasons and developing the first assays and everything else with the grant money, we needed five years. But the process from starting the AI system and screening with the screening assays, it took just a year to generate the scaffold, which had to be optimized in an arduous procedure, traditional way with grant money. And that was the reason why it took longer than it could have been done.
Dr Sarah Cheeseman: But all the same, that's faster than the way things have historically been done.
Dr Marius Galyan: Yeah, that's true.
Dr Sarah Cheeseman: So that's pretty fast. I mean, reaction I'm having when you say that Galyan Bio was started in 2015 is how much you've accomplished in a short span of time, really in the lifespan of discovering and testing compounds for critical uses. It's very impressive.
Dr Marius Galyan: Thank you.
Dr Sarah Cheeseman: I'd love for you to tell us a bit more about Galyan Bio, about its inception, and also a follow on question- we know that you've just you've just joined us in the United States, moving your company here from Germany. We'd love to hear your thoughts on and why you've done that and what you hope, where you hope to go now that you're here?
Dr Marius Galyan: Yes. The company started with first my interest, my personal interest, because in my family, there are several cases of Alzheimer's disease, and as a clinician, I could not help my relatives, and that was the initial idea to do something about it. And it started the whole process and then selected Huntington's Disease because we have good animal models, and I wanted to expand that for Huntington's and for Alzheimer's and Parkinson's disease. And and we initially wanted to really discover new compounds, but we saw that in our assays that these hunting compounds that also were effective in lowering and improving neuronal survival of other diseases. So we have data for Alzheimer's and Parkinson's disease, and we then saw also that the compounds could improve also the survival of normal neurons if we stress them with heat shock. And that was the initial idea to go into longevity and see if these compounds are also useful to for healthy aging. As with aging, our ability to reduce pathological proteins is reduced, so we are accumulating, then we age, dysfunctional proteins, which always come to existence. But with aging, we our ability to remove them is decreased and these compounds are improving that. And that is, I thought the initially that could be useful also for longevity. And that was how we came up with that. And there are several reasons to relocate the company to the U.S.: there's a huge talent pool for drug development in the US, and that's very important for us and also our most important academic collaborators are located here in the US. And and our hunting disease compound is at the moment currently tested by a U.S. company for potential licensing, so that's also a reason to relocate. And last but not least, of course, the funding availability for life science startups in the U.S. is better than in Germany. These were, I think, a lot of good reasons to relocate from Germany to the U.S..
Dr Sarah Cheeseman: Hmm. It all makes sense. Painting that picture. It's a big, big shift, though, from from where you've been. And I know we were chatting before we started recording this, that you're still figuring out where you're going to settle, but you've got some great options, both on the West Coast, and on the East Coast. So we'll look forward to hearing where you land and how you get set up here to keep moving this important work that you're doing. Well, we've run out of time. As I said, it goes quickly like that, and it was delightful to learn more about your background, particularly about the AI piece. I find that so fascinating. And we've really enjoyed working with you and look forward to hearing where you go from here. So thank you again Marius for taking the time today, and we'll definitely be following along.
Dr Marius Galyan: Thank you also for being here. Thank you.
Dr Sarah Cheeseman: Well, thanks everybody for tuning in. We'll look forward to seeing you next time on our next edition of 17 Minutes of Science.