Welcome to BIOTECH NATION !!! With understandable interviews requiring no background in science, BTN attracts a wide global audience. From everyday people looking for hope in treatments in development, to bioentrepreneurs interested in the experience of their fellow travelers, to venture capitalists looking for possibilities in cutting-edge breakthroughs, to scientists simply interested in the work of others, BioTech Nation is the voice of human endeavor, driving science to new realities for everyone. These interviews are drawn directly from the public radio program, "Tech Nation", which also can be heard in numerous global radio and podcasting venues.
Instead of trying to treat just one symptom or two of a disease, instead of focusing on one tiny target on a cell to introduce a new medicine, what if we try to treat the whole disease? Meet doctor Avak Kahvejian, the founding CEO and now member of the board of Cellarity. Doctor Kahvejian, welcome to the program.
Dr. Avak Kahvejian:Thank you. Thank you for having me.
Dr. Moira Gunn:Now most people expect that we're all going to have new drugs come along every day, but you're very clear at Cellarity, drug discovery is on a whole new path and from there, a whole new trajectory. How so?
Dr. Avak Kahvejian:Well, traditionally, we've relied on serendipity or chance to find medicines, and the entire industry is obviously working to find cures for a wide array of diseases. But it's done often through trial and error and painstaking attempts at finding something that works. Back in the day, we used to look at natural compounds from nature to see if they could help. We then moved to making chemical changes and using a variety of chemicals and attempting many, many trials to see if we could find solutions to to diseases. And that has, of course, gotten better and better, more systematic.
Dr. Avak Kahvejian:But at Cellarity, we thought there could be a better way, one where we could use computation and a more wholesome understanding of biology to get to a drug more directly, get to an answer more directly, and increase the probability of success in clinical trials and ultimately bring drugs to patients more readily.
Dr. Moira Gunn:And what is that approach?
Dr. Avak Kahvejian:The approach involves really taking a step back, and instead of going after finding the one target inside your body that we wanna build the drug against, looking at the biology as a whole, looking at, let's say, a cell, Cells are the fundamental unit of life in our bodies. They carry out all the important functions, and they they replicate. They they reproduce. They sense and respond. And in disease, their behaviors go awry.
Dr. Avak Kahvejian:They, do untoward things. And we can see that if we read out their profile of expression, of transcriptomic expression. That transcriptome is essentially a manifestation of which genes have been turned on in that cell and which genes are off in that cell. But when you're in a state of health or in a state of disease, that profile changes. And there are a plethora of changes.
Dr. Avak Kahvejian:We're talking about hundreds of changes going on inside a cell at that level. And by learning and reading that out and having a computer interpret that, we we resist the temptation to look for a single flawed piece in the cell. We instead look at the full picture, compare and contrast healthy and diseased cells, and that informs what types of drugs are best suited to reverse the disease back to health. And, again, that's done using a computational technology that we've built in house.
Dr. Moira Gunn:Now aren't the cells in my body, while they all have my DNA, aren't they different in different parts in my body? Like, my skin cell would be different from, say, a cell in my heart, and hence have to be modeled differently?
Dr. Avak Kahvejian:That's a great question. The DNA in virtually all of our cells is quite identical. You can take a skin cell and a brain cell and a liver cell from one individual, and it will look very, very similar. What changes from cell to cell in order to create the various beautiful functions we have in our diverse organs is which of the genes are turned on and which of the genes are turned off in the various cells. That's what makes a cell a cell, and that's what makes a neuron a neuron or a liver cell a liver cell.
Dr. Avak Kahvejian:So, indeed, there are very large differences across the human body and also between health and disease. There are plenty of changes with respect to which genes are turned on and turned off. So for every disease, we have to study the right cell type or cell types, because indeed that profile, the transcriptome changes between the the various cell types.
Dr. Moira Gunn:Here's my second question. Since I'm a different person from you and have different DNA in my cell profile, my transcriptome, it's different from yours and different from everybody else's. Do we have to look at the particular person or can we somehow generalize?
Dr. Avak Kahvejian:Another great question. At the end of the day, we're all human. At the end of the day, a liver is a liver. Indeed, there are minor genetic differences across the populations. So what we do at Celerity is we often look at many examples from many individuals and focus on the similarities as opposed to focusing on the differences.
Dr. Avak Kahvejian:We wanna find medicines that can work for the broadest, number of people, not a medicine that works for only one person, of course. And so that's how we address that diversity that you're speaking of.
Dr. Moira Gunn:We'll just take cancer. Let's say we know what your healthy cell looks like. You must have some healthy cells. But, again, if you have cancer, some of those cells have changed. We call them cancer cells very specifically.
Dr. Moira Gunn:If we know we have the cancer cell and we know we have your healthy cell, are we trying to drive cancer cell types to healthy cells?
Dr. Avak Kahvejian:Indeed. In the cancer case, that's when not only the transcriptome or that cellular profile has gone astray, We also have that issue at the level of the DNA. So in a cancer cell, there's also changes at the level of the blueprint, which many a time are so far gone, you often wanna just eliminate the cancer and not necessarily try to correct all those changes and bring it back to health. What we've been focusing on primarily at Celerity are diseases where the genome is the same, the DNA hasn't changed, but the cell has been pushed into a different state or is now misbehaving, let's say, and has a profile that has gone awry at the level of the transcriptome. So the genes are the same.
Dr. Avak Kahvejian:It's which genes are turned on and which genes are turned off that's different. And we interpret that those differences using a computer. And now the next stage, which becomes really, really interesting, is that we have a database of many, many, many medicines and their corresponding effects on different cell types. So So we have a database of those changes digitized, and now that gives us predictions. And then we go and test those predictions in the real world.
Dr. Moira Gunn:Now one of the things that becomes clear to me that we often don't talk about is that, hey, I I've got a condition. I'm getting a pill. It's gonna change this. Great. Is that it's changing all kinds of things in ourselves, enough of which are right and enough of which don't hurt us otherwise that it works.
Dr. Moira Gunn:It seems like we understand a lot more or capable of understanding a lot more about how we really work at the cellular level.
Dr. Avak Kahvejian:Absolutely. I think the old traditional way was understand the one potential thing that's broken inside the cell that's not working properly or that's working too actively and design a medicine to engage with that thing. And you can turn it off or you can activate it, and you you're satisfied that you've succeeded in doing that, and then you move forward and try to test the medicine in patients and realize, well, that didn't work. It didn't work well enough. It didn't have the multitude of effects that I really needed to change.
Dr. Avak Kahvejian:And that's why we figured that approach, albeit has been successful, it isn't sufficient. And we can now with computers and with richer information that we can extract from from cells and tissues. We can make better predictions, of of efficacy, of it working, and we can also make predictions of any untoward effects as you're mentioning. We can start to see, is this having effects that we would like to avoid and not use that molecule, not use that that, chemical that will go into a pill, and use another one instead that doesn't have those changes.
Dr. Moira Gunn:Now you mentioned that you have digitized so many drugs that we have, so many of these small molecules as we call them, but also pills, tablets. That's what everyday people call them versus people who are scientists or in the industry. Yet there are lots more that have never been discovered yet. How are you going about that creating brand new drugs that don't exist?
Dr. Avak Kahvejian:Yeah. Great question. So these pills, as you call them, are actually chemicals. They have chemical structures, and our database has a diversity of these. We can learn from the ones that are working and the ones that aren't working.
Dr. Avak Kahvejian:What are some of the shapes and forms that are successful, and which ones aren't having an effect? That will inspire the generation of novel molecules, bringing components of the of the molecules that are working together to create new types of, chemicals, which will become the new pills of the future. And we have scientists who can look at the data, and it'll inspire them and inform them as to what steps they should take next. It's in some way a recommendation engine, if you will. It sees the things that are working, and that informs what would be likely to work next.
Dr. Avak Kahvejian:And that's how the new new, pills emerge.
Dr. Moira Gunn:Of course, we always like our models to get smarter, and some of those recommendations may work and some may not. How do you take these recommendations, suggestions, new ideas coming out of what's possible and actually train your model to get better? Where's the ground truth as we often say?
Dr. Avak Kahvejian:Yeah. That's why it's so important to have a digitally native organization and to have a company that's focused on collecting information along the the journey in digital form and feeding that back into the the algorithm, so that the computer gets smarter and our models get better. Because we want to see the experimental results, and those real world experimental results will inform whether we were right or wrong, as you mentioned. And we want that to be fed back into the system so that going forward in the future, the next round of recommendations take advantage of those lessons. And that's exactly what we do at Celerity.
Dr. Avak Kahvejian:So the real world experiments have are recorded. The molecular data is collected from those experiments and fed back into the algorithm, and it improves its predictive power every time we use it, every time we we go through the cycle.
Dr. Moira Gunn:And I have to tell people if they don't know, the reason we keep saying chemicals is that when we find a biological drug of any sort and why is it small molecule, if it's small enough, then we can chemically synthesize it and put it in a pill. But so many of our drugs are not, such as for rheumatoid arthritis. They're huge. And so there's not small enough to do. So not only is it small enough we can make it into a pill, into chemistry, we can use chemistry then, the chemical elements in our model, which is really terrific.
Dr. Avak Kahvejian:Yes. Absolutely. The building the the chemical building blocks can be recombined and redesigned, pretty much at will. So we can design very customized, molecules to put into a pill format. The the other interesting thing about pills is that you swallow them.
Dr. Avak Kahvejian:So you can create oral drugs as opposed to having injectables, and that's another big advantage of of what we're doing. Traditionally, it's been very hard to discover small molecule pill based medicines. As I said, it's been done through trial and error, attempting 1,000 or millions of of compounds and looking at which one will work. Here, we're creating a much smaller set, thanks to the informed recommendations from the computer, and we're refining it and creating a new entity. But at the end of the day, the destination will be a pill.
Dr. Avak Kahvejian:And that's desirable in in many indications or many diseases where needles are not desirable, where you want folks to take the medicine at home, or you wanna deploy the medicine worldwide, in places where health care systems can't support injectable therapies, let alone more complicated therapies, that that hospitals can barely administer.
Dr. Moira Gunn:Now right now, Celerity is working in several specific areas. Conditions of the blood, immune diseases, and with Novo Nordisk, NASH, the liver condition which some people know under its former name, NASH. Let's focus on 1, you know, perhaps, the blood conditions. What are you looking at there?
Dr. Avak Kahvejian:Absolutely. So there, we're looking at, one disease in particular we're very excited about is sickle cell disease. This is a condition where you have a mutation in the gene that makes hemoglobin, and that results in hemoglobin having a funky shape, and causing the your red blood cells, which are normally nice plump discs, to take the form of a sickle, and that's where the disease's name comes from. And that in turn causes your red blood cells to get caught in your small blood vessels. It's a very painful, effect, and it it causes what's called vaso occlusive crises.
Dr. Avak Kahvejian:And the disease really doesn't have any good cures to date. One extreme approach is to go all the way into your bone marrow and into your DNA and to correct that mutation that's causing hemoglobin to fold incorrectly. But what we've done is looked for small molecules or or or something that can be a pill that can change the behavior of your red cells such that they use an alternative form of hemoglobin, what's known as fetal hemoglobin, carries the same function as adult hemoglobin. It can bind to oxygen and and, resolve all your symptoms if you make more of it and change the balance inside your red cells. Reduce the amount of mutated hemoglobin being made and increase the amount of fetal hemoglobin being made, and thereby eliminating, that misfolding, eliminating the sickling in in the bloodstream.
Dr. Avak Kahvejian:And it'll be in the form of a pill. So you can swallow it, and it'll impact your your, your bone marrow and correct the the, the deficiency. So that's something we're very excited about, and our platform has has, has already borne fruit, and we're progressing that forward.
Dr. Moira Gunn:Now, obviously, this is only the beginning. You're not just sitting there saying, no. This is all we're doing. This is all we're doing. Where does this go from here?
Dr. Moira Gunn:Originally, I said, well, you can this is a new way to discover drugs, but also there's a different trajectory here. How do you see that trajectory?
Dr. Avak Kahvejian:Well, hopefully, it, you you could see that because we're thinking about such a generalizable, approach, anywhere we can relate disease to cellular dysfunction, which is many places, you named a few, immune disorders, diseases of the liver, such as MASH, anywhere we can relate those diseases to cellular changes and the changes of the cell's transcriptome or its profile, we can most likely find medicines to alter that profile and bring it back to a state of health. So that's why we're so excited that this can be a new way of discovering medicines, one that is more holistic, more deliberate and directed, and, computationally powered, which will result in a higher likelihood of success as we go into patients, and and, deploy the the medicines.
Dr. Moira Gunn:So how the heck did we get through this interview and you never said the term artificial intelligence?
Dr. Avak Kahvejian:Well
Dr. Moira Gunn:I love it. I love it. But there's artificial intelligence here. Yeah.
Dr. Avak Kahvejian:Yeah. Absolutely. We we we we don't wanna pepper this with, with, you know, buzzwords, but, indeed, I think the the computational algorithms we're using are all based on, artificial intelligence and and machine learning. Machine learning allows us to interpret patterns within very complex sets of data, and these transcriptomic cellular profile changes are very complex, and they're very dynamic. And one human brain can't capture all of these changes and keep them in in in your head.
Dr. Avak Kahvejian:So so without machine learning, this would not be possible.
Dr. Moira Gunn:Well, I have to thank you very much for that because I'm on a mission to say artificial intelligence is not magic. It is it is it is really programming and data and being very deliberate about how one does things. And, so I wanna thank you for that. And, doctor Kavejian, you're always welcome, here, and so please come back. See us again.
Dr. Avak Kahvejian:Thank you. Thank you for the opportunity. I think we're very excited by what we're doing and happy to share it, with the world.
Dr. Moira Gunn:Doctor Avak Kahvejian is the founding CEO and now member of the board of Cellarity. More information is available on the web at cellarity.com. That's cel, sell, a r I t y, celerity.com.