Dr. Barkha Yadav-Samudrala: 0:10

Hello, everyone, my name is Barkha and welcome to nerd RX podcast. And today we have a very interesting topic called high throughput screening. And to talk more about it, we have one of my best friend, Dr. strugglin. Mandala. Welcome shovelin to the show.

Dr. Shravan Morla: 0:29

Thank you, Barkha. It's a pleasure. Thanks for inviting me.

Dr. Barkha Yadav-Samudrala: 0:33

Yes, of course. So let's get into the topic right away. Why don't you tell our listeners something about your background? And how did you come to this point where you are now?

Dr. Shravan Morla: 0:47

Sure. So I have a pharmacy degree from India and India, you can actually get a pharmacy degree during bachelor's so I had a chance to work in a pharmacy lab and explore different areas of pharmacy in terms of medicinal chemistry, pharmaceutics formulations, I spent each semester in a different setting, trying to understand what what kind of projects that each lab does, and what does that specialization mean. And I fell in love with medicinal chemistry, which is basically trying to design drugs, you think chemistry to design medicines, that is what Mastering Chemistry is right. And I decided to get a PhD in medicinal chemistry and I joined Virginia Commonwealth University with you. For my PhD, and I'm getting my PhD, I realize that some mathematical chemistry also has different subtypes, right, you can be in a medicinal chemistry lab making compounds. Or you can use computational tools to screen these compounds. Or you can use biologicals tools to screen the compounds or by chemists. And I tried to tried my hands in all of these and and I realized that testing these drugs in a biological setting is what motivates me the most is one my cup of tears. And that's what I've been doing since and that is what we're going to talk about today, high throughput screening. And as well, right now, I am A High Throughput Screening scientist at a biotech called Generate biomedicines, where we generate biomedicines. and test them using our high throughput screening system, which I'm very excited to talk about.

Dr. Barkha Yadav-Samudrala: 2:40

That is amazing. And by the way, we both started our PhD together and ended our PhD together, and it was a fun ride down. But a fun one. Yeah,

Dr. Shravan Morla: 2:52

definitely a lot of memories and downs like everyone faces during their PhD program. But it was worth it.

Dr. Barkha Yadav-Samudrala: 3:02

It was worth it. Like as I think I mentioned in my introductory episode, like keep at it. And then I think for me, it was more downs than ups. But in the end like today, when I look back, I'm like, thank goodness, I did not quit. So now I'm like, it has definitely for me, but I'm pretty sure for you, as well has like you have become a better scientist because of all the doubts especially.

Dr. Shravan Morla: 3:30

Everything's a teaching moment for sure. Yes, exactly. For some of our friends, maybe stopping where that had to stop was probably a random session. And just for the listeners, if you're going through this, especially during a PhD process, we don't have so much help. So there are people out there you can talk to seek help from whether it's emotional, mental issues or anything. Just wanted to let you know that we are all bent to the same thing. And we are here to help.

Dr. Barkha Yadav-Samudrala: 4:03

Yes, and let me also emphasize that nothing is more important than your mental health. So if you ever feel that you're drained, just take a break, talk to someone who is in your situation or a professional. And trust me there are like a lot of people who are like, ready to help because they have been through it. That's right. Like I think I have realized through my PhD that we are scared maybe to talk about this stuff, especially like mental health issues. But once you talk to the right people, they are so welcoming. And so they're like they're very helpful people out there.

Dr. Shravan Morla: 4:45

We don't realize how common this is. Till we talk about it.

Dr. Barkha Yadav-Samudrala: 4:50

Yes, yes. And so let's talk about mental health as well. Because if I think people focus more on the physical well being but I think the Mental health is equally important. Yeah. Yep. So let's get into high throughput screening. So let me start with what is basically high throughput screening. And what about it interests you the most?

Dr. Shravan Morla: 5:19

It's a very difficult question to answer straight away. Because High Throughput Screening is not a technique. It's okay. It's a process. It basically takes any technique that you use, whether it is a radio ligand binding assay, a calcium influx assay, or an enzyme activity assay, any of these assays that you usually use to screen your compounds, let's say a 96, well plate or one compound at a time, right, and then convert this into a more rapid and robust process. So high throughput screening, as the name sounds, is basically screening a large library of compounds, making it a very high throughput. So automating these processes is what high throughput screening does. So at my previous job, we screened a library of 400,000 compounds in a matter of two to three weeks, using this high throughput screening, combat screening setup. And if you read and since my company was more of a smaller biotech, our libraries, our size is limited to 400,000. But if you go to larger pharma, like Pfizer, AstraZeneca, I was reading a paper from AstraZeneca the other day, and they're their internal library size is 1.3 million compounds.

Dr. Barkha Yadav-Samudrala: 6:41

I thought 400,000 is huge.

Dr. Shravan Morla: 6:45

Yeah, so yeah. 400,000 is probably not even a quarter of 1.3 million. Yeah. Yeah, so. Yes. So it's basically, as you know, drug discovery is a very time taking process, right? Yeah. And depending on the protein, or the target, or the receptor you want to hit, right, whether you're finding an inhibitor, or activator, agonists, or antagonists, you usually need a starting point, right, of how to design your compound of how to hit this target. And in most of the cases, there's no starting point, right? And even most of the successful cases from earlier days are mostly found through serendipity. Right? You cannot just look at the receptor and say, All right, this is a binding pocket I think will fit the best there. It is. Not possible, right? So yeah, the starting point. And there are multiple ways to find these starting points. And High Throughput Screening is one of the one of the most widely used technique, it's, that is something I like about it, because you are starting with the unknown. You don't even you're screening tons and tons of compounds. And you will find, hopefully, find a head that is very specific to your target. So you are the starting point. And that's what I like about it, I am initiating something, I'm initiating the process of drug discovery for this target, and I get the sense of ownership and starting something and that excites me.

Dr. Barkha Yadav-Samudrala: 8:28

Okay, awesome. Yeah. And I think with high throughput screening, the it also makes the process a lot faster as compared to like you mentioned, we just look at the receptor target, and then decide what fits there. So it's like generating a hit or a lead compound there. So can you just quickly go through the process? Like what is involved in high throughput screening?

Dr. Shravan Morla: 8:57

Sure. So the first thing for high throughput screening is obviously, depending on the project, I'm assuming that you already have a target, right? You need to know what protein what GPCR, what enzyme you're targeting. Right? And that is how you have a program or a pipeline, if it's a company, right? You have, let's say, a company working on trying to target Kira, which is an oncology target, right. So you already have you already know that all right, I want to work with Karis, right? So you need to have that protein made in huge amounts. Right. And for that, you have protein scientists you can make the protein in different ways. So exploiting and purifying the protein would probably the first part if you're doing a purified system based high throughput screening or if you want to do a screening in cellular systems you need to know what type of cell does. If a good says cell type, to type the indication you Want to test? Right? So having that starting point is important. And second thing would be to validate that starting point, let's say you're working with this kid, as I mentioned about, are you really sure that this character is responsible for cancer? Is it in some way a validated target, right. And that's very important. Why? Because you don't want to spend years or months in a hyper screening campaign, and it requires 1000s of dollars, right? And so many people's time, so many scientists go into it, and then find out that inhibiting Kara's actually does not do anything in cancer settings, right. So it is very important to validate the target. And the third thing, and, and I think is, what I really liked the best about hybrid screening is to develop a screening assay. Right, you want to develop an assay that captures what this protein or the target does, you need to have a method of testing the activity of this target or the function of this target. And this is usually called asset development. This process is called assay development. And it is very important that the assay you develop is physiologically relevant. And the reason why it is important is because you might be able to test the activity of this function of this protein in multiple different ways. But eventually, what your assay is reading is the function of this protein, right, or the binding of this port into the compound. If that does not correlate to sell your settings and in vivo settings, all your high throughput screening campaign is not useful, then you want the heads from behind CT screening campaign to eventually correlate to the cellular settings or the in vivo settings. Right. So having a really good screening assay is important. And the next step would be optimizing this assay for A High Throughput Screening format. And what I mean by that is that when you're usually developing an assay, whether in an academic lab or an industry, you develop it in a 96. Well, plate, right. And the volumes in 96. Well, plates are typically about 50 microliters, usually 100 microliters, right? And that consumes a lot of reagents. Yes, and when you're testing, let's say with AstraZeneca, 1.3 million compounds, that's a lot of creations right. And some of the assays we use the kits we get use a reagent for detection. This is one of the assay I was showing the other day a kit which comes which can be used for 10 plates, Justin plates of 384 Well, plate costs about 13,000 Oh my goodness, right. And you if you're doing 1.3 million compounds, it just adds up real quick. Yeah, especially if you're doing it and 96 well plate that's just not feasible, right. So you want to manage many miniaturised this assay as much as possible. So people tend to do either 384 well plate or in most companies nowadays are trying to do 215 36 Well plates. Right? So we said

Dr. Barkha Yadav-Samudrala: 13:17

1536? That's right. I think the biggest wallplate I have ever used is like 96. So I'm just imagining what does 1536 looks like,

Dr. Shravan Morla: 13:28

you cannot actually see through it. Like you can see through it. But it's it's like very difficult like.

Dr. Barkha Yadav-Samudrala: 13:38

So do you like a manually do it? Or

Dr. Shravan Morla: 13:41

No its all automated. But for 384? You can manually do it. I wouldn't recommend it. If you're doing more plates. If it's one plate, maybe you can do it. But 1536 I haven't seen anyone manually use it. It's all automated. Yeah. And that's, that's going to be later step on the High Throughput Screening cascade process. Right? So yeah, so people spent a good amount of time transferring this 96 Will assay to 384 or something 36 Well format, based on the resources they have, right? And the bigger number of well you go, the smaller your assay volume would be. Because if you look at it, the 96 Well, plate size is the same as that of 1536 well plate size. So the plates dimensions are the same, just that the well that might have been volumes smaller and smaller as the volume becomes smaller and smaller, right. And in 1536, you probably only use like four to five microliters and assay volume coming from 100, microliters and 96. So you reduce your cost or free or reduced all your reagent cost, right. So optimizing this is also important because sometimes scaling down is not easy. Yes, you kind of lose your segue The assay window, all of that. And once that is optimized, the, you have your system ready. And then the other part of hydro screening begins where it which involves compound libraries. Right. As I mentioned, different companies have different sets of compound libraries. And some companies don't have their own compound libraries if they're smaller combat companies or recent biotechs. And for that reason, we have multiple CROs who have millions of compound libraries that you can actually even transfer your assay to their CRO, and they are going to screen the library for you and give you the data.

Dr. Barkha Yadav-Samudrala: 15:38

Okay, so even that step is taken care of like, that was my next question. Like, if you have like million of compounds, like the library from a CRO, how do you pick which library to go for?

Dr. Shravan Morla: 15:51

Exactly right. So the thing with high throughput screening is the don't know what we're looking for. Right? We don't know the compound structure and the large number of compounds you screen, the better chances of finding a head, right? It is usually a needle in the haystack. And the more haystack screen, the more needles you'll find, right? Yeah. So yes, so, we have these, what are called as Asha ready plates. That is this, these plates already come with pre dispensed compounds. Wow. So that you have this plate with compounds in it and in their field. All you do is open the seal, and then transfer your agents to it. And it's so cool that we have this at our company, we have this Hydra system. It's from this company called Hydra solutions, which integrates everything that is it takes the asset or the plate from the incubator. There's a robotic arm, which takes it from the incubator opens the sale, puts it in a liquid handler, which dispenses your reagents. And then from there, it does all the let's say you want to incubate it at 27 degrees Celsius, it puts it in an incubator, which is at 27 degrees Celsius. After the certain time, wherever you want to do it, it takes it out. If you want to wash the plate, it washes a plate, right. And then add whatever you want to add it next, and then read set, and then give the data. And this is really useful one for multiple reasons because of this integrated automatic robotic system. You are constantly working the system, as scientists can probably do 2030 plays a day. Right? Because you're you're dispensing that one instrument, you're gonna walk to the second instrument reader incubated a third instrument, you have to take a lot of manpower into it. But it integrates everything in one area. The robotic arm kind of moves it from one machine to another machine. And it can run 24/7

Dr. Barkha Yadav-Samudrala: 18:05

I am so jealous. You have a robotic arm to even like open the seal of a plate. Yeah, oh my goodness, just

Dr. Shravan Morla: 18:12

feel and feel the bag, which is so cool and satisfying to watch.

Barkha Yadav-Samudrala: 18:15

Oh my god, like that's what like the other day I was just doing these BCA assays for my Western Blot. I have like 96 well plate. But this is tiring. And of course, we don't have a robotic. So I do manual additions to all the wells. And at times, I was like, did I add to this? Well, like it's so confusing,

Dr. Shravan Morla: 18:35

especially as the number of plates increase the chances of error also increase, right? Especially and with the volumes decreasing. There's no way to actually see and say whether you added this to micrometer reagent or not. Right. And with a robotic system, you don't have any of those problems as long as your program is set up. Right. And you have a very user friendly system to even set up those programs. Yeah, you don't even have to know programming.

Dr. Barkha Yadav-Samudrala: 19:05

Yeah, like I remember my PhD advisor would say like when it comes down to a human and a machine that machine is always right.

Dr. Shravan Morla: 19:14

Well, he's probably right. They don't take over the world but

Dr. Barkha Yadav-Samudrala: 19:20

yeah, yeah, that's the hope.

Dr. Shravan Morla: 19:23

Yeah. So yeah. So it just speeds up the process right. Yeah. And then you get this data, which again is for me, I screen 400,000 compounds which was probably like seven 803 84 well plates right. And that gives you so much data, right look, you there is no way you can look at individual compound and say okay, this compound is doing bad compound is good. You need data analysis packages, which look at multiple setup at a single time. And there are multiple of such packages which are people Solving every day, but then kind of give you with what we call as hits. hits are compounds, which do what you would want for a drug to do. And then that is your starting point for your drug discovery campaign. So hydropool screening, kind of the whole process gives you a starting point to start with.

Dr. Barkha Yadav-Samudrala: 20:23

Okay, so just I know it will be different for all the libraries. So for example, you're screening these 400,000 compounds, how much of the compounds do you think would actually be hits?

Dr. Shravan Morla: 20:37

That is a great question, right? The usual hit rate is in a high throughput screening campaign, at least in my experience, I've seen multiple, a couple of projects, I've worked on a couple of projects, my colleagues worked on this usually around 0.5%. Wow. Right. So if you're screening 400,000 compounds, that would be how much

Dr. Barkha Yadav-Samudrala: 21:07

less to calculate that's like

Dr. Shravan Morla: 21:09

2000 compounds, approximately, that's still huge, right? That's still a lot of compounds. So then you have to sit down with these 2000 compounds, and then understand whether these 2000 compounds are actual hits. Or they're what we call positive false positives. Right? Exactly. And even if they're not false positives, they could be doing something else, right. And this class of compounds called paints, which, if you have, which we ever heard of, in our, in our classes with Dr. Glenn, his talk about them, it's these compounds, which show up in almost every at every screen, right? So you have to be careful about these compounds, right? You need to, you need to have a good eye to look at them or experienced chemists to point out that okay, this this compound has upgraded in multiple screening campaigns, it's probably not a good starting point for us to start with, you don't want to go with that compound and end up doing all your Exploratory Studies, because, again, it will go nowhere.

Dr. Barkha Yadav-Samudrala: 22:21

So is there also a library for paints,

Dr. Shravan Morla: 22:24

there is there are multiple resources to look for pain compounds. And, of course, as with many of these data, some of them is publicly available. And many big pharma have their own libraries. Because as you test more screens, as you do, as you have a bigger pipeline that if you're testing multiple targets, you can you start building up your own libraries inside your own data packages, right. So there are some publicly available pain libraries. And some come with experience and some come with the privilege of working in Big Pharma, I guess. Yeah.

Dr. Barkha Yadav-Samudrala: 23:09

Okay. So now that you have 2000 hits, probably, I would say 1500, because 500 are pains, sure, what is the next step,

Dr. Shravan Morla: 23:19

the next step. So hydropower screening is usually done at a single concentration. That is you're testing the ability of this compound, let's say at 10 micromolar, to inhibit this target. Right. So that is just whether or not attend micromolar this compound index is target. The next step would be to confirm the screening, because you're doing one point screen. Right. And as you know, we usually want it in replicates, right? So believe in your data, right? So there are multiple counter screens, depending on the system. The thing from here is, there's no there's no set path, every person, every company can do it in different ways. Right? You can do what we call a dose response screen. Right? So now you have like an eight point or 11 Point dose response instead of a single point, right? And then you understand from that, you get an EC 50 or an IC 50 value. Right? And one way of looking at this data, is that all right, I want compounds with the best ice 50s. Right? You have you kind of reduce them. And the other way is looking at counter screening, let's say you have a very similar target, which you don't want to hit based on the homology of these two proteins. I want to hit Protein A but not protein B. So you would count screen all your heads with protein B and anything that is hitting both protein A and protein B would be in the trash. Yes, right. Then you further Uh, bring down your compounds, so a little bit

Dr. Barkha Yadav-Samudrala: 25:02

less. So like you, you mean, you want to make selective compounds for that one particular target

Dr. Shravan Morla: 25:08

exactly right. And then you'll also do a few counter screens, which are usually well established within each company, right? We have an assay setup to find aggregators, right? Small molecules, especially because of their very hydrophobic nature tend to like form this aggregate, right, the right number flick system, we don't. And the assets readout can be influenced when they're forming this aggregates. So we don't want to work with those aggregating compounds. And there's a way to figure out whether these compounds are aggregating or not. So you do those assays. And there's weed out those compounds also. And depending on the assay compounds library you're working with, and the protein you're working with, you have multiple other counter screens, like you have a counter screen to find out redox compounds if you don't want redox compounds in your assay, right? And then you'll eventually go down to like 400, or 300 compounds or something. So you may bring it down, bring it down, bring it down to a comfortable point where you think, all right, I can now look at each and individual structure. And then we'll put them into compound classes, right? Because let's say we have 400 final hits, all these 400 compounds are not going to be individual compounds. They're going to be analogs. Right. Right. So you would put them into chemical classes. And most of all of this is work one on one work done by chemists. So hybrid screening kind of involves multiple different people from different backgrounds, we have protein scientists to make proteins, we have a biochemistry by a physicist to make these assets assay development, right. And then you have automation scientists who to automate this entire process in this hydro system, for example. And then you have chemists coming in to look at your data.

Dr. Barkha Yadav-Samudrala: 26:57

So you have jobs for everyone,

Dr. Shravan Morla: 26:59

we have jobs for everyone. That is right. Yeah, so then chemists would come and put them into different classes, and then they'd be like, alright, this come this class of compound is very difficult to work with. And then if you look at the analog data, all of them seem to have pretty similar ice 50s. So the SAR is pretty flat, will not be able to work with that, let's just put them on the shelf. Right. And then the second class, these are all very fatty, they have like 10 alkyl groups on them, not probably good compounds to work with put them aside, right. So you'd probably end up with like three or four families that you're comfortable working with. And then the chemists would do all the SAR explanation around them doing the heat rate optimization. That is

Dr. Barkha Yadav-Samudrala: 27:41

another topic for another

Dr. Shravan Morla: 27:47

show that is what A High Throughput Screening screening campaign involves all the way from deciding your target to finding ahead.

Dr. Barkha Yadav-Samudrala: 27:57

So from 400,000, you kind of bring it down to two or three classes of compounds, guys, right. So how long does this process usually take?

Dr. Shravan Morla: 28:09

It depends, right? It depends

Dr. Barkha Yadav-Samudrala: 28:12

on the assay development, and it depends

Dr. Shravan Morla: 28:15

on multiple different things. Let's say you have your protein, your assay, everything ready. And we're only talking about screening, right because heightened screening kind of put screening at the center stage. So if you're only screening, what 400,000 compounds for example, it would take if everything is automated about three to four weeks. Okay, yeah, probably even less but in my just

Dr. Barkha Yadav-Samudrala: 28:48

being conservative Yeah, that's that's not a bad timeline.

Dr. Shravan Morla: 28:52

Yeah, and in if you're doing it by hand, you will probably only strain that compound by that.

Dr. Barkha Yadav-Samudrala: 28:58

Definitely. Yeah. Yeah. Like I know you like by the sound of it. It looks like you love high throughput screening. But let me play the devil's advocate here and ask you what you don't like about it? Is there anything you don't like about it?

Dr. Shravan Morla: 29:17

I enjoy the acid development Parsh portion the most okay, because that is very thrilling to me in terms of developing something trying to you know, understand the biology and then bring it to in vitro system you want as I said, your assay should correlate the physiological setting as much as possible right. That takes into account a lot of you know, you are more putting in your brains into assay development. And for me High Throughput Screening once automated does not it involve any of your own energy and Right, exactly, you don't need to put in brain for that. Once it's automated, right? Yeah, I'm trying to give as much appreciation to the automation process. It can be, it's not easy. And we need persons with the skill set right to do that. And that is very interesting as well. But the robot doing a thing is really nice One day One day, but after that, it's pretty boring, you know? Yeah.

Dr. Barkha Yadav-Samudrala: 30:34

But I think if you think about it, it's like the necessary evil. You're like, you need the robots to screen through such huge libraries. And also, like you have, you don't have like the human errors, like you mentioned about pipetting, like, two microliters into each well, like, we can't even see that. So I think it's like necessary at this point.

Dr. Shravan Morla: 30:56

I agree, right. And if I had to pick between doing it manually, or automating it would automate any day. But that can be a little bit boring for scientists sometimes. Right? I would say that's not a downside. But that's how the system works, right. But what I don't like about it on a more serious note is that we are exhausting our compound libraries. Right? We are screening, like, if you look at the trends in terms of hydrogen screening, in the initial days, let's say 1900 9090, all of those screening campaign gave more hits, which actually reach the clinic, compared to screening campaigns these days. Because we are screening the same libraries with a different target. So we are exhausting our libraries, right. And that's not so much unique in terms of race coming every day. So the success rate is going down from hyperthreading. That being said, it's not impossible, right? People are still doing high throughput screening on a daily basis, people are finding so many hits on a daily basis, which are rushed in the clinic. But the success rate is going down, in part because of lack of novelty in high throughput screen in the screen libraries. Because screening libraries, as I just mentioned, we need millions of compounds. So what happens is that you have let's say, a benzene scale or something, right? You make minor modifications to it to an existing compound to make a different compound and increase your library size. But are there very big variations and scaffold in the complexity as a few strengths? All of those, I think those reduce that's one thing. And the second downside is, sometimes it's very difficult not to realize that you're dealing with or you're working with false positive, let me say

Dr. Barkha Yadav-Samudrala: 33:08

it, right. Those pains.

Dr. Shravan Morla: 33:12

They might or might not be pain, but or it might not be a false positive also, but sometimes it's you, you realize you're spending a few years before knowing that okay, this is not doing what I'm what I think it should do. Okay, and that is a more that's a story more common than you would think also. Right? So you have to be very careful about working with discount bags sometimes.

Dr. Barkha Yadav-Samudrala: 33:39

Okay. So in terms of, I know, like you covered what you guys do in the industry. But what does High Throughput Screening looks in academic settings? Because I'm pretty sure no lab is spending$13,000 on a kid to do like, just 10 kids out there. And obviously, I have not seen a robotic arm in any labs yet.

Dr. Shravan Morla: 34:05

So high throughput screening is a very, what do you mean by high right? Is 400,000 High is 1.3 million high? Right? So that level of high ranges from every place and it depends on where you are right? I would say academia, or regular academia mostly does what I like to call medium throughput screening. Their compound libraries are usually less than 10,000 or so. Right? And in a more conservative setting like VCU, where we have been the compound libraries are more than less than 1000 compounds. Right. But it is people are building up libraries. People are purchasing libraries from external sources and people are trying to you know don't do as much as screening as possible. Like, I can just give based on my own experience at VCU that you don't necessarily need a robotic arm, specifically in let's say, 5000 compounds, right? It's a few 384 well plates or even if you're doing in 96, well plates for 5000 compounds, you can probably able to do it just fine. Right? And people do get good data out of it, good results out of it, good compounds that can be translated into the animal models. I'm not sure if how successful academic labs happened to translate into the clinic. And there are many success stories I don't want to say no. And, for example, I can tell you about my experience at Scripps as a postdoc, and we had a unit within Scripps, where they house a lot of chemical libraries. And they used to bring in these chemical libraries, and I'm pretty sure I might be wrong, but their chemical libraries are in the five to six figure compounds. So close to 100,000 compounds, I'm probably more right. So academic. I wouldn't say academic labs, but academic institutions are like hubs, right? Like in San Diego, we have Scripps salt, UCSD, Stanford, Burnham, like a couple of different institutions, which pool their resources and form this collaborative academic Union, where they can house and libraries from different vendors, different sources, and like what Dr. Desai was trying to initially do at VCU, build a screening facility for ISP 3d, right, instead of structural biology, drug design, and development. So more and more academic universities are doing that. Whether an academic lab by itself can do it, it just depends on the funding available and how important that screening campaign is to that project. But an in depth of the second question of whether academic lab can afford a $13,000 worth of kid? Well, that's a good question. And even for industry, it can be difficult sometimes, right? With lot of funding. There's ways around, right. So to do screen, one particular target, there are multiple different assets that you can develop, right? The asset I was talking about, probably gives you a better asset window is more robust and everything right, but then it's also very costly. But then there are nine other different assays, which are less expensive. But you might have to compromise on your asset window a little bit, or you might have to be, it's less robust or something like that. Right. So academic labs can find these different solutions to it.

Dr. Barkha Yadav-Samudrala: 38:03

There are like ways for everyone to do high throughput screening. Exactly. It doesn't have to be. You can Yeah. And as you mentioned, I think collaboration is key, like I think more Institute come together, I think there are better chances to like have this screening facilities and everyone benefits from it.

Dr. Shravan Morla: 38:22

Exactly. Right. Yep. To the for the greater good anyway,

Dr. Barkha Yadav-Samudrala: 38:28

for the greater good of science and for humanity, for humanity.

Dr. Shravan Morla: 38:32

That's why we are all in it.

Dr. Barkha Yadav-Samudrala: 38:34

Yes, yeah. That's the goal. So like, as you mentioned, I know we diverted a little like, when you have those like you have screened and eliminated like most of the compounds, you come down to those two or three classes of compounds, if I'm not mistaken, they're called leads, right?

Dr. Shravan Morla: 38:54

Well, the STL hits, okay, we are just classified into different families. And then you would work on those families of heads to slowly change them to a lead compound, a lead compound is something that will be ready to put in an animal.

Dr. Barkha Yadav-Samudrala: 39:13

Okay, again, like ready for a preclinical phase

Dr. Shravan Morla: 39:17

Exactly. With the head. So the head to lead optimization would be starting from the families to going to maybe one or two lead compounds, you don't usually have more than those lead compounds, right. And in this process, you'll be testing it in different models, like you'll probably go away from a purified system and you'll probably be tested in different cells, different cell lines, different biophysical states, right and different vessels, like analyzing them in different contexts, like not just a phenotypic cell screen, but also endpoint based cellular screen that And then looking at your selectivity and specificity. And also, is this compound developer developable in terms of ADME properties, right? You might not necessarily test it in an animal in terms of aadmi, but you have theoretical models and predictors of, you know, adding properties. So, all of these are taken into consideration in your hit lead development, and also a couple of things like, Does this seem to affect compounds? Do they affect important activities? Like, do they block the potassium channel in the heart? Right? Or do they have any effects on the serotonin receptor? All of these a few well established markers, which you don't want to hit necessarily, are usually tested to make sure that you're not hitting them. In terms of safety.

Dr. Barkha Yadav-Samudrala: 40:53

Yeah. Okay. Yeah. So like, I think from what we have learned today, it's like taking this huge library and cutting down to like, what one or two compounds that is right, and from all the stories going on, there are again, chances that might not work in the field. Right?

Dr. Shravan Morla: 41:12

Right. So starting off with your high throughput screen doing to the lead compound, it might take one or two years or even more, depending on, you know, the success rate. And once you put in an animal, right, that is when you would know if it's working or not, right. And when I mentioned that, a lot of times, it's too late to realize, like, it's just, you spent so much time, that's what I meant, because your cell heads, or your head from all of these screens don't necessarily translate in vivo. And you have to go back and see, okay, maybe this was wrong, and it's just very disappointing that.

Dr. Barkha Yadav-Samudrala: 41:53

So it is like a continuous process. So, for example, if you end up in this scenario, like it did not work, would you like, again, go back and screen a new library for the same target?

Dr. Shravan Morla: 42:07

That that's a really good question. It depends, right? It depends on a lot of different situations, financial situation being one, right? Whether the company thinks it's a good target to pursue whether. And when, when you go back and analyze your screen and your data, and you realize, maybe your entire process was good, just that your lead optimization probably was not right. You know, if that's the case, you will probably like just change one of your leads. And then that's a different model. And sometimes it is possible that the animal model you're testing itself is wrong. Okay, right. So you might just change the animal model. Right? And sometimes, you realize that everything was right, it's just that protein is difficult to head. And then you might not want to pursue that. So it's very situation based and finance base and necessity base.

Dr. Barkha Yadav-Samudrala: 43:14

Yeah. Wow. So I think just reaching to that one lead compound takes couple of years. And then you can, like, we all know, like from preclinical to getting it approved is like whole another battle people face and I think definitely, we should cover that in the future.

Dr. Shravan Morla: 43:34

And working with animals, you should probably no animal work takes a lot of time.

Dr. Barkha Yadav-Samudrala: 43:42

Tell me about it. Like, the experiments, which I'm like working on the animal behavior. Like, the animals are so finicky, like just today, they will be like, okay, they will perform the experiment 100% I'm happy and tomorrow. They're like, No, I'm not going to do it. And I'm like, What do I do? You know, that is one thing I used to love about chemistry. Because when I make a compound, I have things like NMR mass spec, and I can proudly say yes, I'm 100% sure that I have made this compound. But with animal research, it's always in the gray area, you know, so that is something I definitely miss about chemistry.

Dr. Shravan Morla: 44:24

Yeah, yes. You are dictated by how that reduces. Yes. Experiment it depends on whether the cells are ready or the animals are ready, right? Yep, yeah.

Dr. Barkha Yadav-Samudrala: 44:39

Well, I think this was a very fun episode. Thank you so much, Shravan, for all the insight and please let me know if you have some reading material for our listeners. I will make sure and put it down in the description. And once again, thank you so much, Shravan, for being here.

Dr. Shravan Morla: 44:58

Thank you. Thank you so much, Barkha for inviting me it's been a real pleasure catching up and I wish you all the best of your podcast.

Dr. Barkha Yadav-Samudrala: 45:06

Thank you so much Robin and listeners. I will catch you next week on another fun episode. And remember, it's not a bad thing to be a nerd by