- 2 years ago
BullFrog AI is an artificial intelligence company that has developed AI technology it calls bfLEAP. The platform is designed to transform the field of precision medicine by applying predictive analytics to all stages of drug development. In doing so the company aims to improve the lives of every patient.
The company has just announced exciting data from a preclinical trial for a therapy targeting glioblastoma.
The company has just announced exciting data from a preclinical trial for a therapy targeting glioblastoma.
Category
đź—ž
NewsTranscript
00:00 All right, Mr. Singh, how are you doing today?
00:06 Great, Aaron.
00:07 How are you?
00:08 I'm good.
00:09 Welcome to Benzinga's All Access.
00:11 Thank you for taking time out of your busy Friday to join us.
00:14 Let's hop right into it.
00:15 Why don't you just give us an overview of Bullfrog AI to start?
00:20 So Bullfrog AI is a tech-enabled drug development company.
00:25 We have a proprietary AI that comes out of the Johns Hopkins University Applied Physics
00:29 Lab and we're applying it to the challenges in drug development with an objective of reducing
00:36 the time and investment, increasing success rates in this long and high-risk process.
00:44 Got it.
00:45 I mean, it sounds like, I mean, I know the people are at Johns Hopkins.
00:47 Those are some smart folks.
00:49 So it sounds like you guys are working with some good people.
00:54 You say you aim to improve lives by identifying critical connections between therapies and
01:00 highly responsive patients in less time, with less cost and more precision.
01:06 How are you guys doing this?
01:08 So with AI, you can find patterns, relationships, anomalies that you can't find using any other
01:17 technology.
01:19 With our particular approach, we use unsupervised machine learning and graph analytics.
01:25 And the advantage there is it allows us to analyze what we call shallow and wide data
01:30 sets.
01:31 A lot of times, for example, in clinical trials, you may have a few hundred patients.
01:37 And traditionally you would say that's not a lot of data.
01:40 That's not true big data, but it's a wide data set because you have genomic data and
01:44 things like that.
01:46 So even looking at data sets like that, we can identify information and take a precision
01:54 medicine approach to drug development through the efforts.
02:00 Got it.
02:01 And so Vin, tell us about what makes the Bullfrog Leap Tech different from other available technologies
02:08 in the space.
02:09 Sure.
02:10 This is a technology, like I said, that was developed at the Applied Physics Lab.
02:14 It's a subsidiary of Johns Hopkins.
02:17 They have more than 6,000 engineers and scientists there developing all kinds of amazing technologies.
02:23 They do a lot of defense related work.
02:26 This technology won innovation of the year there years ago.
02:32 And we've continued to license improvements to the technology.
02:36 They had a publication that came out over a year ago comparing our proprietary algorithms
02:42 head to head with the open source algorithms that we all hear about.
02:45 These are the libraries available through Microsoft, Amazon, and so forth, Google.
02:51 And our algorithms in that situation outperformed these algorithms for speed and accuracy of
02:58 prediction.
02:59 And then just a couple of months ago, APL submitted an application to the R&D 100 Awards,
03:06 which is the global Oscars for R&D, and the technology was a finalist in those awards.
03:13 So that's a pretty big deal.
03:15 That's recognition.
03:16 So we're definitely in what we'll call the innovator category for AI.
03:20 We're not applied AI.
03:23 Applied AI is just using these open source algorithms.
03:26 And we continue to innovate.
03:29 Got it.
03:30 And then so, Vin, I know you guys kind of recently started trading on the NASDAQ.
03:36 Can you tell me a little bit about your decision to IPO and what that process has been like?
03:41 Well, we had options for sure.
03:45 We thought the timing was right.
03:47 AI was starting to really heat up and made the decision that the public market route
03:55 is the best route for us because our company has tremendous potential.
04:02 We have a unique business model.
04:05 We think there's a lot of attractive elements to what we're doing that made us ideally suited
04:11 to be a public company.
04:13 Got it.
04:14 Yeah, I mean, I'm sure it's been a busy time for you all for sure.
04:19 This spring, I mean, I know you've mentioned it a few times now, but this spring you partnered
04:22 with Johns Hopkins.
04:24 What has that relationship been like?
04:26 Why did they choose to harness your technology?
04:30 So Johns Hopkins, we have multiple relationships with them.
04:33 Like I said, our AI comes through their subsidiary.
04:38 We've also licensed multiple drugs from Johns Hopkins.
04:42 Our two glioblastoma drugs were licensed from them.
04:46 And I think Hopkins, they believe in what we're doing, they trust us, and they think
04:53 we're going to add a lot of value and make a difference.
04:56 And ultimately, I think these institutions are more focused on just making an impact
05:01 in society, and I think they know that we have the tools to do that.
05:06 Yeah, well, like I said, I know they've got some great people and some very smart people
05:10 over there as well.
05:13 So Vin, how has the development of Bullfrog 222 or BF222 progressed?
05:20 So 222 and 223, they're both glioblastoma drugs.
05:26 They're part of our program.
05:28 We're looking for a strategic partner to advance these programs.
05:33 And once we identify that partner, together we'll put a strategy together about how best
05:39 to advance these two drugs, or we may decide just to advance one of them.
05:45 We had some interesting animal data that came out on 223 recently.
05:50 There was a press release that came out where we demonstrated that both of the drugs are
05:56 essentially equivalent in terms of their impact on glioblastoma.
06:02 However, there's been earlier studies with 223 that showed it has improved bioavailability
06:08 and solubility.
06:10 So more of the drug can get where it needs to go, and you can take less of that drug,
06:15 basically, is what that says.
06:18 So we have an interesting opportunity here with these two drugs.
06:24 And we'll decide as soon as we secure this partnership as far as what we want to do going
06:30 forward.
06:31 Got it.
06:32 And then, so backing away from Bullfrog specifically and talking more just generally about AI,
06:40 I mean, this has obviously come burst on the scene this year with Chad GPT and is now in
06:46 the public eye.
06:48 How rapidly has the field of AI advanced in the last few years, and where do you see it
06:52 going in the next few years?
06:54 Well, it's really exploded.
06:57 I think consumer-facing AI through Chad GPT and so forth, it's unbelievable, the growth
07:06 rate.
07:07 I think companies now are providing other companies with tools, AI tools, to help them
07:14 improve their business processes.
07:18 I see this continuing to accelerate, continuing to find new applications.
07:27 But at the end of the day, it's going to come down to results.
07:31 It's great to have this tool, in some cases a toy, but it has to add value and produce
07:39 results.
07:40 And I think you're going to see more and more of that, a lot of proof points, especially
07:44 looking at healthcare, which is our space.
07:48 I'm expecting the number of applications where a true difference is being made is going to
07:54 increase probably exponentially.
07:57 And I think we're going to be right in the middle of that, because we're tackling these
08:02 problems from a critical point, which is the drug development process.
08:09 Ultimately, we want to improve success rates in drug development.
08:13 Right now, big pharma fails about 50% of the time in phase three, which is unbelievable
08:18 considering they spend $1 billion to $2 billion in 10 to 15 years to develop these drugs,
08:23 and they have all that experience.
08:26 There are a lot of patients that could benefit from these drugs if they were developed in
08:31 a way that leveraged true, powerful AI, like bullfrogs, to dial in the right patient, right
08:41 drug, right dose.
08:44 And that's the direction we're going.
08:45 So in healthcare, I see precision medicine having a lot of success going forward.
08:52 That's what I'm expecting.
08:54 Got it.
08:55 And Ben, I mean, your guys' business is very much in the science field as well.
09:02 What is the key to marrying science and business together?
09:07 I think the key is having the right team.
09:12 We have a lean team and a very diverse team, but a very experienced team.
09:19 Most of the people on my team have 25 plus years of experience, and they come from diverse
09:24 backgrounds.
09:25 You have people who come from biotech, pure drug developers.
09:30 You have people from academia.
09:32 You have people who come from leading companies like Takeda and Thermo Fisher, and they know
09:39 how to build a business.
09:41 So it can't just be about having great technology.
09:45 It comes down to execution.
09:48 It comes down to under-promising and over-delivering.
09:52 And when you have that right combination, that chemistry to do that, and you operate
09:57 with urgency, great things can happen.
10:00 And that's how you have success.
10:03 And in our industry, it's very important to do that.
10:07 We've been very smart about how we've stepped forward with our business as it continues
10:13 to develop and mature, learning from the mistakes of other companies, trying to do things the
10:20 right way, being very smart with our cash.
10:25 And that's pretty much it.
10:26 That's the formula.
10:27 Well, yeah, I mean, it sounds like you guys are on the right track, Vin.
10:31 I mean, it's been, you know, like Mitch said, we've talked about this company on the show
10:36 before I traded it.
10:37 So it's been awesome to learn more about what you guys are doing.
10:41 So like I mentioned, AI has been the hottest story of 2023.
10:48 How similar is your technology to the chat GPT stuff that's out there that's kind of
10:53 capturing the attention of the general public?
10:56 And then how far do we need to go to reach a true artificial intelligence?
11:01 Yeah.
11:02 So chat GPT is very different.
11:05 It's scraping information from the web.
11:08 OK, first of all, we operate in what's called a closed system.
11:12 OK, so we don't have to have concerns about, you know, information that's, you know, inaccurate
11:23 or opinion, you know, things like that.
11:26 It's information.
11:27 It's information that comes from our client to us.
11:29 We analyze the data and send it back to them.
11:32 I mean, that's I think that's a key differentiator for us.
11:36 Obviously, we're very focused on health care only.
11:40 We're not focused at all on other industries.
11:44 But and then a couple of other qualities, I think, that separate us.
11:49 Number one, we can make predictions even with incomplete data, which is a problem in every
11:53 sector and in our sector.
11:56 It's no different.
11:58 We work with multimodal data so we can literally bring any type of data, genomic data, imaging
12:03 data, lab data, physiological data, so on and so forth.
12:07 We bring it together and we can build these networks, graph networks that we can then
12:12 query to answer questions.
12:15 And then finally, the biggest differentiator is explainability.
12:19 OK, our AI is explainable.
12:23 That GPT, at least to my knowledge at this point, is not.
12:27 You don't know where it's getting all of this information.
12:30 And some of it, you probably don't want the information, but it's in there.
12:34 With ours, we can actually show how we arrived at the answers, how we've made the predictions.
12:42 And it's our belief that if a scientist or physician is going to make life or death decisions
12:48 with a AI tool like ours or make a billion dollar decision about whether to pursue a
12:54 drug, it needs to be explainable.
12:56 It can't be a magic black box that spits out an answer.
13:00 This is really important stuff we're doing.
13:02 So it's critical to have that capability.
13:05 As far as a true AI, I think we're probably a ways off from that because I think a true
13:14 AI will need to have intuition, emotion, opinion.
13:20 And I think you're probably looking at 15 to 20 years away before we have reached that
13:27 point.
13:28 Yeah, I mean, it sounds like that kind of true AGI, the artificial general intelligence
13:35 that people that the type of stuff you see in like sci-fi movies still seems like it's
13:39 pretty far away.
13:40 And I mean, we'll see.
13:42 But for now, people are still pretty enamored with the chat GPT version.
13:47 So who knows what it'll look like once we get the more advanced technologies?
13:51 Well, Vin, like I said, it's been great to learn more about Bullfrog AI today.
13:57 Again, the ticker is up on the screen.
13:59 BFRG now trading on the NASDAQ.
14:01 Vin, thank you for hopping on Benzinga's All Access.
14:04 Hopefully we'll get a chance to reconnect in the future.
14:07 Anytime you guys have important company updates, news events, et cetera, I would love to have
14:11 you back on.
14:12 Great, Aaron.
14:13 Thanks.
14:14 It was great to be here.
14:15 Have a great day.
14:16 Of course.
14:16 (upbeat music)
14:19 (upbeat music)
Comments