Precision Cancer Medicine
In our deep dive, we discuss how detecting, diagnosing, and treating various disease states should be viewed as multi-dimensional and encompassing all aspects of human genomics, including genomics, transcriptomics, proteomics, and metabolomics. By integrating these diverse layers of information, we can gain a more accurate and holistic view of the disease, leading to better-targeted and personalized treatments for cancer patients.
Furthermore, we delve into the rapidly growing field of bioinformatics and its crucial role in aggregating, integrating, and analyzing these vast datasets. We’ll explore the latest tools, techniques, and computational methods being developed to make sense of the wealth of data generated by multi-omic studies and the challenges and opportunities associated with managing and interpreting this information.
Join us as we examine the potential of precision medicine in revolutionizing cancer care by offering more targeted and personalized treatment options, reducing side effects, and ultimately improving patient outcomes. We’ll also discuss the broader implications of these advances for the healthcare industry, including their potential impact on drug development, clinical trials, and healthcare delivery. Don’t miss this in-depth analysis of the exciting future of precision oncology and how it’s set to transform cancer care as we know it.
One such macro trend is multiomics approach to precision cancer medicine.
We’ll specifically be looking into precision medicine with a focus on cancer.
We will dive into bioinformatics as well, and we will go through our agenda and finish up with the analysis. A few process comments, so we are not soliciting any investment or giving investment advice in any way whatsoever.
This presentation is general industry research based on publicly available information.
The new asset you try to keep this information private, this is only for in internal use and not for the general public.
We have invited you to this because you’re a technologist, thought leaders, entrepreneurs, industry experts, early adopter customers or sophisticated investors that are part of the I Select network.
We value your thoughts, questions, comments, and insights into this topic and we greatly appreciate it if you actively engage during the presentation. Thank you in advance for your attendance and active participation.
So I would kindly ask that you all please go on mute. However, if you have questions, please feel free to unmute and jump in at at any time.
You can use the chat window to ask you a question as well. After the formal part of the presentation, we will open your microphone So you can ask questions or offer advice. Anyways, again, feel free to unmute. And then there will be time for questions discussion at the end. And last but not least, this presentation is being recorded and will be available for replay.
So today, we’re gonna talk about a multi omic multiomics approach to precision cancer medicine.
I just wanna caveat this real quick.
I do not have a science background. I do not have a cancer background. However, my younger brother had retinoblastoma when he was one year old one years old.
So this is a personal topic for me, and it’ll be something that he’ll have to deal with for the rest of his life.
So the topics being covered today. First, we’re going to look at the big picture.
Then we’re going to dive into sort of the background by addressing some definitions.
Some motivation behind this. We’ll discuss big data a little bit.
We’ll look into a potential solution or as well as early studies.
And then we’ll dive into the market. We’ll look at things from a global perspective, things from an investment perspective.
We’ll also look at a couple of market trends that that that I’ve identified.
While it’s not all inclusive, those are the big topics that I felt more important to pass, and then the global informatics market.
After that, we’ll dive into the who and what, talk about the latest and greatest some some companies that are really making some changes in in big strides in this industry, leading investors and exits, which there are not many in this topic right now. Then we’ll look a little bit into the business model, specifically the monetization of the data. And then we’ll finish up as we accommodate with the analysis.
And then that’s when I’ll really welcome your thought and comments and insight.
So the big picture here, So everybody’s aware of multiomics studies, but they’re really the keys to understanding our complex human biological system.
Produce a massive amount of data.
But as we all know, that data is not information.
We need to be able to figure out how to take that data and translate it into meaningful information to draw inferences and insights.
Gonna look at bioinformatics as it’s a new scientific relatively new scientific approach to data collection storage and analysis.
Really is the computer science way of manipulating large scale data for precision medicine.
So development of enabling technologies, one thing we’ll discuss today will uncover mysteries in healthcare.
But then it’s also good to think about what it’ll what it can translate to beyond health care.
And then we’ll kind of look into some of the major players.
So we’ve identified Google, Microsoft, IBM, Apple, no surprise, have increased expenditures as of late into AI specifically targeting healthcare.
And then should we believe the hype?
There’s a lot of buzz around artificial intelligence. There’s a lot of buzz around machine learning. Computer science lately, especially since Google released or there was a CB Insights article describing how Google was gonna use AI to tackle this IBMs, Watson, Apple’s sort of consolidation of electronic medical records and Microsoft’s investment.
So there’s a lot of hype, a lot of buzz, but should we believe it?
So we’ll talk about that too much.
Everybody’s familiar with precision medicine.
But what this allows us to do is target individuals specifically rather than take an average and utilize a one size fits all for that average approach to diagnosing as well as employing and implementing a plan for medication and so on and so forth.
A multiomics approach going to discuss this a little bit. So this is really the integration of all the OMEX. There’s a significant amount of data that has the potential to be uncovered, while it is not uncovered, yet we are making strides to do this. There are several institutions that are taking a multiomics approach.
There is a lab out at Stanford School of Medicine that is really focusing on this, in particular.
And then more specifically within cancer diagnostic six as well.
So the omics studies by their nature rely on large numbers, significant massive amounts of numbers of comparisons and tailored statistical analyses.
So now that we’ve identified what this multiomics approach is, what are some of the advantages?
There’s a lot of words here. I trust that you all can really read and digest, so I won’t go verbatim.
But the big thing is we’re taking one dimensional studies that have limited information to multidimensional studies. To really understand the etiology of various diseases.
There’s a lot of information out there, so the integration of this.
Human genomes, as we know, are all complex, multiple levels manifested by various genomic assays.
And then again, you’re gonna utilize integration for these approaches.
But the bottom line here, as you’ll see our little friend in the bottom right corner, it’s probably a little confused.
Big takeaway here is big data meets machine learning, meets complex biological systems.
So what is bioinformatics?
Some of you all may have heard of it, may have looked into a little bit.
There have been a few TED talks on this topic.
But much of this information has been coming out in the past couple of years.
MIT in in Harvard, have developed, as well as Boston University, have developed labs to really start studying bioinformatics.
But really what it is is a hybrid science that links biological data with techniques of information storage, retrieval, distribution, and analysis to support multiple areas of scientific research, including biomedicine.
So mining these data sets that these scientists are able to uncover will improve, for example, microbiome analysis.
Accelerate immunotherapy development and really help us understand the etiology of disease. There are a significant number of other advantages for bioinformatics that we can go into and discuss at a later time.
There is a new newer development called translational bioinformatics.
In all honesty, I’m not one hundred percent certain on what the real difference is.
As you’ll see, and I’m sure you all have seen in your studies, there’s a lot of overlap in some of these definitions, in some of these subsets or subgroups of biology.
So feel free again, hop in if you have anything you want to add.
Here’s a little clipping from a article As you can see, May twenty third two thousand eight. And just on a side note, you’ll notice that there’s a lot of articles, a lot of clippings that I’ve put into this presentation.
That are from within the past thirty, sixty, ninety days.
So this is really an emerging you know, technology and development here.
So institute Curry is located in France. I apologize if I pronounce that wrong.
But really, they’re looking into high performance computing, utilizing artificial intelligence to accelerate genome sequencing and inter rotation for oncology.
So why is this important?
Because we are seeing partnerships. There’s a large amount of data We’re seeing partnerships with large technological firms such as Intel, such as Microsoft, such as Google, and so on, that are really looking to partner with small labs, small institutions to take this to the next level of disease etiology, prognosis, diagnostics, and so on.
So what’s real motivation here?
As you know, there’s a lot of hype between AI and machine learning, but more specifically, high throughput sequencing.
These are significant levels of data that we have discussed.
And when we start taking a multi omic approach instead of a single omic approach.
These data databases, data sets start to increase exponentially.
And we really have to learn on or rely upon machine learning.
So there’s been a buzz like I said, for an accumulation of this data, we’ll go into discussing big data here in a little bit.
But is it overwhelming? We don’t know. But people want more data, especially a lot of these big technological firms, as well as some of these large scale pharma companies.
So with this data, there’s an urgency to develop these powerful tools to efficiently and effectively search this information, search this I’m sorry, search this data and filter this data to gain insights and inferences.
And then, obviously, there are some resources that have come out such as Athena Geniasis, as well as NetICS mostly open source little to no cost databases, as well as some other databases from large scale institutions that we’ll look at here in a minute.
The existing tools that we have are fragmented.
So we’ll discuss that as well.
So this gentleman here is John PA LowAnidis, And as you can see, he has a very science heavy background.
But he specifically focuses in biomedical data science and statistics.
He’s from Stanford University.
And pretty much so he’s not a necessarily a skeptic.
Of this, but he wants people to understand that just because you have large sets of data doesn’t make you smarter.
And so that’s kind of an issue that we run into.
Sometimes it may be overwhelming, and we may start going down the wrong wrong holes. This is Jenna Wiens.
She’s the assistant professor of computer science and engineering at the University of Michigan. What’s interesting is that I’ve had a few conversations with some people on the West Coast.
We won’t name them specifically, but they’re very excited about the Midwest.
Very excited about capabilities at some of these institutions, such as Michigan, Indiana, Ohio State, northwestern, as well as Washington, some of these capabilities that they have. So instead of really looking to the coast, we can start to really cultivate this and and engage with thought leaders from the Midwest.
Okay. So we have an interest in this. Now we need to collect inform collect this data.
So we’ll go through a couple of examples here about of how this is being conducted.
NIH just recently, and as you’ll see May first, open a nationwide enrollment for a huge precision medicine initiative.
So they’re looking to get a million people to sign on to this program and divulge all their information to be able to make advancements into biology, more specifically precision medicine.
So there’s a great path ahead.
Should this really take off one million participants is a little daunting. So let’s look at another potential solution.
This is project baseline. This is being sponsored by Google, Verily, Duke, and Stanford.
So the initiative of this is to collect information, or I’m sorry, collect data from ten thousand people based upon various demographics.
And to really understand some of the healthy aspects and some of the some of the more unhealthy aspects to be able to draw insight and inferences.
From their data that they are divulging.
And then we have more specifically the cancer genome Atlas.
Now, you all may be familiar with this to a certain degree.
I cannot speak too intelligently upon this, but, essentially, what it is, they’ve compiled data from eleven thousand tumors from thirty three. I’ve heard thirty two gone back to thirty two, thirty three to move point of the most prevalent forms of cancer. But more specifically, they’re looking at pan cancer.
So not just targeting an individual cancer, but across entire cancer system.
As you can see on the right side, there is an upcoming conference. I am not being paid by them.
I’m just trying to pass some valuable information.
So we’re collecting this data. Now what?
Is there a solution?
Unfortunately, there’s not quite a solution just yet, but we are on a pathway to really accelerate this PPM, what we call predictive for better and personalized medicine.
We have made advancements and strides into cancer systems biology, as well as computational biology.
And then we have the bioinformatics approach, and then going into translational bioinformatics as well.
So these are all approaches that are being taken to speed up this process.
A couple of early studies, a lot of words here.
Don’t expect me to read this off.
But I’ve highlighted and bolded the more important parts.
Integrated analysis of genomic and transcriptomic data revealed potentially important targeted therapeutic response related events and proposed new molecular classification of breast cancer patients.
And then also this integral analysis transcriptomic and proteomic data. Not deciphered, that was not able to be deciphered through single omic data sets.
Was utilized to better understand complex complex cancers.
And then multi multi omics approach has provided a large number of potential biomarkers and targets.
And it will still unfortunately take a long time to fulfill the long term benefits, such as sensitive early diagnosis, prognostics and and etiology.
But we’re looking to significantly improve overall survival.
And then the last one, a quantitative multiscale imaging of the breast in breast cancer is an emerging field. And these were all pulled from the documents set in the bottom right hand corner.
And again, with that, if there are any any references or or points that you would like me to pass along from some of these journals and reports and whatnot.
Please feel free to reach out either in this presentation or you can email me my email addresses at the end.
So now we’re going to talk about the market First, taking a step back looking at global precision medicine market.
So I’ve seen several statistics about this. There are numerous reports, none of which I have the money to afford as they all are over five thousand dollars.
But really, the big picture here is this industry, this market has the potential.
It’s expected to reach a value of over one hundred and forty billion dollars by year two twenty two thousand and twenty six.
But more importantly, the oncology market has compound annual growth rate of ten point four percent or ten point four percent over a nine year span.
So looking at personalized medicine, if we look at the health healthcare organizations where they’re looking to focus their personalized medicine programs.
You’ll see diabetes and common cancers, as well as neurological diseases.
What’s really important here And this would require very additional thorough research on this part to really identify and explore wide diseases, aging, longevity, and diabetes have their interest more so over over common cancers.
Taking a look from the venture side, you’ll notice that since twenty fifteen, we have surpassed the two billion dollars investment mark And that number two billion increased to a little over two point nine billion between two thousand and fifteen and twenty seventeen.
I think we can attribute this to a lot of the large scale players becoming involved in this as well as the public information of these large scale pharmaceutical companies pushing to drive it.
They’re around and, you know, if they’re and the insanity around immunotherapy as well.
This is not all inclusive, as I mentioned earlier, These are just some of the very large scale market trends.
There’s been a lot of talk about electronic medical records. And here, I’ve information was pulled from Warton in a report that I can share later on.
Really, what this is, we have the Internet of things. We have this electronic capability.
Why can we not put this on electronic medical records?
I won’t go into this too much because I know you all or most of y’all are are very familiar with this.
But one thing of note that we’ll touch on it in a little bit. Apple has made strides in this category.
To align with thirteen medical institutions around the country to have access to this type of data.
We will look into privacy information sharing and the regulatory environment.
Genetic analysis is taking off as we’ve talked about. More importantly, it’s a linchpin of precision medicine.
So partnerships like we have discussed, large scale players, working with some of or early stage companies in taking on their resources to be able to support this advancement to precision medicine.
And then there are obviously trends in big data and AI, and there’s a growing trend in bioinformatics as well.
But more importantly here, the advent of precision medicine would entail changes to every facet of modern healthcare.
So let’s dive down a little bit deeper and look at the global informatics market. As you can see on the bottom left, twenty thirteen, it was roughly two point five billion dollars to three billion dollars for the overall market.
By twenty twenty, it’s expected or forecasted to grow to twelve point eight six.
Now if you shift your attention to the bottom of the screen, this is from a little little over a year ago.
But this industry is expected to grow by twenty twenty five to ninety-five point five billion dollars that’s a pretty big shift.
It’s a pretty big change.
And that’s something that should definitely be be inspected and really evaluated.
Why is it expected to to grow that much in seven years?
And then the three other columns to the right won’t dive too deep into this, but this is some other information regarding the bioinformatics market. So the who and the what who are the players?
Who are the companies? What are they doing? And so it begins.
These, like I said, The dates on these are all very recently.
I won’t go through each and every one of them, but I’ll highlight the important ones.
Aspire Ventures launches a three million dollars fund for tech supporting precision medicine.
Google focuses on healthcare data, Bessmer launches a seed fund for startups, Microsoft.
They not to say that they are just getting in.
They have been in this as well as IBM.
But what we’re seeing here is a lot of information from these large scale players, a lot of money and a lot of momentum being pushed into this.
Twenty fifteen, Barack Obama, start of the precision medicine initiative. I know that there’s been some skepticism with the current administration without getting into politics.
It’s my understanding.
There have been no significant changes, especially with the recent NIH proposal for this study of one million individuals.
So some of the early stage companies I just kinda wanna caveat this real quick.
In some conversations that I’ve had with some individuals within this industry, A lot of these companies are kind of in the shadows, which may seem odd, but there’s a specific reason for it.
There’s a hype cycle around technological advancements.
And so there’s kind of been this I guess, common belief that within this field, they’re not trying to push the valuations of early stage companies.
To highly inflated levels.
So they’re taking a a very precise, very precise approach to how they’re raising money.
A lot of these companies have alternate names or aliases that they operate under. Non descripts I’m sorry.
They don’t publish widely what they’re doing, what they’re working on, but they are actively seeking and actively targeting venture investors, institutional investors strategic investors beyond the light of publicity.
January is an interesting company because it’s more of a B2C play.
So they’re aggregating this data and being able to develop and provide a platform for the individual consumer to monitor, to track their health pattern.
But what data are they aggregating? I see two sets of data out here.
You have real medical records that come from your position.
Right. And then you have things like genetic testing, wearable device information, self reported information. Right.
So when we talk about these guys, are they aggregating medical records into this?
So there’s three things. There’s the biological data.
There is environmental data as well as lifestyle.
So do you drink more than two beers on the weekend?
Do you live in an area like Fukushima, for example, things like that. They’re taking into consideration That’s self reported.
That’s self reported information then. Right. And then you can do the genetic testing there.
I’m guessing they’re not using healthcare records and, like, the information from your doctor.
So, there are some companies that are that are doing that, that are partnering with clinics.
And obviously, there’s a consent that there has to be an authorization to to to really provide this data from the individual patient level.
But to my knowledge, people aren’t actively going out and just pulling information at will.
It’s it’s all being initiated on the patient.
As far as providing this information.
Does that — Okay.
So there are whole It seems like there are a bunch of sort of dot or eight kind of gather data kind of big data as what’s going on. Mhmm.
And have we seen anybody come up with a business model that pays the providers of information for their data, so it could be used.
I mean, it is a dot com model that has — Right. — early profitability.
So we’re gonna get into the business model innovation here — Right.
— in a second But yeah. Obviously, there’s Voice on the phone. Right? What’s up?
Maybe she just went off the phone.
Is she on the phone or she is? She was No. I’m I’m here. I was just muted. It took me a little to get it unmuted.
Melissa, we you should have been at this conference. We were out of WashU. It was really some of the top people really in cancer.
But one of the things that came up was the use of synthetic control arms Have you heard that term?
Synthetic? Yeah. No. Not familiar with that.
So I believe what are you familiar with the I Spide effort at Laura Esterman, U.
S. US side U. C. San Francisco?
Only very vaguely.
Only the name.
So what she’s doing and she’s what she’s doing is Following all the patients in great detail bypassing the HR and getting much more data around breast cancer and she’s got twelve center supporting her.
And we’re really sort of jumping at jumping ahead a little bit.
She discussed some static control arms, which is that she she knows all these various treatments that have been done to patients. And she can FDA apparently starting to assess the concept of if you have patients that have gone through various kinds of treatment you wanna understand performance above baseline for a new drug against other things because they’re starting to get comfortable with the idea of synthetically building the control group out of existing data, and using that to accelerate the FDA process.
There’s a lot packed in there. But what I’m sort of interested in and I could be able to resolve it now.
But I’m interested in is there a business model that allows people to gather data so that drug companies can pay for that data in a way — Right.
Yep. — the FDA is happy. Yep. And it takes drug approval cycle from x to half of x.
So that is a real economic interest in in sort of stirring pot here and speeding things up.
So I’m not sure if there’s a question there for you, Melissa, but that’s sort of a a thing that’s really on my mind at this moment.
A couple of slides, we’ll be able to discuss this a little bit more in-depth.
Another thing that I’ve got in there is is that it just feels like the improved innovation.
We got to improve cycle time. Right. Absolutely.
And there’s all this sort of weird cycle times because of patient subscription, FDA process.
How do we get your own domain faster? Do we get Right.
Or or how do we how do we test hypothesis faster and then go back and understand why they’re wrong.
Against population. How do we not spend hundreds of millions of dollars and and spend a decade researching something that will ultimately fail. How do we know Well, we’ve got that with immunotherapy right now. It’s just this huge amount of money that’s being spent on immunotherapy.
We’ve had some stumble.
And all evaluations are going to get screwed up basically because the hype emerged.
We’ve got how many immunotherapy trials going on. Yeah. Like, there’s ninety.
And I we’re gonna have we’re gonna have seventy failures.
Which is going to really put events in the universe. But keep going, sorry.
No, Carter, that’s – you brought up a really good point.
So with the business model innovation, we’ll kind of it’s not really innovative in the sense, but it’s kind of a slightly new approach. Nothing too far of a stretch in the data community.
With some of these tech giants.
But then also, you talked about how do we get to to to know or how do we get to failure faster?
How do we spend less money? How do we spend less time? How do we spend less resources?
But at the end of the day, how do we save more lives and how do we how do we lengthen the duration that somebody has, you know, really on this earth But else it brings up how do we learn from failure?
So, yeah, these are these are very, very good conversations, very good topics that we need to keep discussing.
But not just discussing amongst ourselves.
We need to really identify who the key leaders are, who the thought leaders are within this, who are the ones that are not just talking about it, but the ones that are doing it, and that want to do it.
And we need to engage them. Okay.
So some of these early stage companies, as you’ll see here, So there’s this is like the distinction that we really need to make.
This is not a traditional science per se.
They’re not going out and executing clinical trials, tests, or studies. These are companies that are working on the capability to really implement computer science large scale, massive sets of data.
And and and and and how do we draw inferences? How do we identify patterns how do we recognize things within this data?
So this is where where I really believe.
And, again, this is my opinion and and my opinion alone.
But I believe that this is the push that some of these large scale players, like the Google’s, the Microsoft’s, and the apples and the other large scale tech giants are really looking to make a make a push.
Obviously, they’ve got the interest of large scale pharmaceutical companies as well. So let’s talk a little bit about some of the leading companies in this, and their financings.
So we’ve got recursion, pharmaceuticals, DNA, nexus, zymurgeon, personal genome diagnostics, Coda, and Tempas. These are some of the large scale players.
As far as, you know, early stage companies.
So the leading companies, they’re generating pretty good amount of investments, a pretty substantial size.
They’re bringing in some serious some serious fundraising. But what’s more important is that they’re getting the attention of some very strategic investors as we have really already discussed.
So I won’t go into all this. You all can can read this on the screen.
Companies are at the top and then the investors obviously at the bottom.
Some of the early stage investors, you know, we have we have a list.
This is not all inclusive Obviously, Bessemer and Foresight as they have launched massive funds to really explore this.
I’m sorry, Bestamer and Inspire Ventures. SoftBank has made a pretty substantial dent into this industry.
Luxe capital, so on and so forth.
We’ve seen a couple institutional investors step into this space.
I suspect that we’ll really start to see a handful more in the really in the coming weeks months and year.
Some of the strategic investors already really discussed these. This has Bill and Melinda Gates.
This has their attention.
They’ve made several grants into companies in this industry. As you know, he’s very passionate about precision medicine.
So I suspect that there will be others as we start really digging into this, that will come out and and really try to make make it an advancement into this So as far as the exits, this was a little difficult, I’ll be honest. It’s been a pretty good deal of time trying to find out as far as an M and A perspective.
But flat iron, obviously, I think you all know as well as next bio you might not be familiar with, but alumina.
Is. So, again, these are companies that are able to take data process. It’s store it.
And then really draw key insight for precision medicine.
So Carter, to your question, how are they monetizing this?
So companies like tempest and flat iron have this Google Esc and there are also a couple other companies out there that are doing this as well.
What they’re doing is they’re providing a platform, they’re providing a database online.
That you can submit data to little, at little or no cost and utilize their algorithms to be able to draw information and and insight.
So where does how do you make money off of this?
I guess if you’re an investor.
Really, it’s on the back end.
So Carrie, you submit a large set of data for analysis.
Mike, you submit a large set of data for analysis, and so on and so forth. And and we’ve got these exponentially massive sets of data.
Who’s interested in this?
So we have Google. Right?
We have large scale pharmaceutical companies.
We have these tech giants that are looking to make an impact into healthcare. And so when they’re able to file all this data, you know, the limited So have a explicit business interest in it or is this just yet one more removal thing?
To me, Google still only makes money on search. Right. And makes money on nothing else.
Now they do other things to sort of drive stuff, but Right. But is they is there a way for them to ultimately monetize it or is it they expressed anything in that regard?
Not to my knowledge, but this is the play.
I have not uncovered anything as far as you know, these pharmaceutical companies or tech companies trying to acquire this data.
But I think if we look at the forecast for the growth within the industry from what is it currently twelve billion dollars to ninety-five point five billion dollars I think that’s where we’ll see some of this interest in some growth. I mean, we’ve seen that if that control arms, if, you know, another trend in the issue is is it As you get more personal lines, it’s harder to build groups for testing and so it’s harder to take longer to subscribe people that it would seem like there’s a CRO like function or CRO like support function — Mhmm. — that basically says we’re gonna go get a million people to relieve us of HIPAA.
Right. Gather all their data. Right? Get all their blood.
We’re gonna get and at least create some kind of longitudinal control group.
And will sell it to FDA will sell it to will be the pitch book of I just wonder whether — Human data.
— because of the nature of data that you’re getting, whether it’s gonna be the data orgs that fill that role or work. The data orgs that fill that role, you know, like the national cancer. The national cancer is to the dot org.
So, like — Oh, dot org. — the nature of Right.
Yeah. But I thought the nature of the information, you know. Maybe.
And maybe they contributed to a regular external model, whether it’s an example of a dot org that has gathered a whole bunch of information.
But it would look at, like, the early days of m p three. So early days, m p three, everybody hated it now.
What are the names of these music?
And they had the metadata from the music and to some degree they could do it, but there was no database of it.
And it started out originally at a whole bunch of dot org for crowd funding and then assembling it and pulling it together.
And then finally, someone actually looked it over in a business model they charge.
It grew really, really fast, and they sold it for a hundred million dollars two guys in it.
But you with that kind of transition that always goes on with these kind of things.
Or and there are zillions of dot orgs and there’s a history of lack of cooperation.
The American standards inside me this won’t talk to the breast cancer inside me.
This research or the one talk to that researcher. Right.
I won’t give you my data. I don’t wanna share my data. And so it’s like this pissy kind of thing.
Things like money to stop.
But is there an example of an effort to get population there any business model in the past to get population data? You know, some great Google who will collect all this information about the So, yeah, there I guess there’s — Yeah.
Good. I mean, thanks.
As well as Facebook.
So Google does have their bare lease subsidiary.
Which is specifically focused on this kind of stuff or it’s one of their main areas of focus.
And then it is in the Google healthcare thing that CB Insights published the other day.
It specifically talks about the fact that they’re going to have to figure out a different business model because to your point everything else they do is on paid advertising, not on any other model. So this would be different and unique for them.
It, you know, it would not shock me to see someone like that acquire one of the EHR systems and start running something in the background on with the data that sits there and exists today.
So, you know, I mean, Epic’s a privately held company.
Well, from that one, because they might do fine.
I think EHR is the data science of Pampers, and I think EHR suck because there’s huge data error and then and and that there’s a huge amount of incorrect and inaccurate reporting just to because the the interface was so bad. And even what the same John’s in push and said that that the the EHRs or the way they’re being forced to write the EHRs.
Is not conducive to clinical.
Even though they the lab results sit in there in a common format, Even that, apparently, no, weird that the other day I’ve heard it before, researchers will not talk to anybody else in touch. So if you go to Quest and get a test on any of these deeper things, but you know CBC or from the old rerun the CBC, they don’t crossed.
It’s very common that they will not cross anybody else’s desk. Now that might be they should.
But surprisingly, they cannot trust other people’s data. You would you would think that the the error if you had a large enough sample, you’d be able to identify where the error was in the in the tests So I still go back I still go back and think that G matters is in a unique position to of someone who’s marrying genetic testing with the EHR data to be able to do plop something on top of it that says they can do some of the some of the analysis particularly as it relates to cancer or other genetic specific, you know, areas. So there there are four rays or or oncology, nytology, neurology.
I think one thing we had to check is if is what is the gauge error in these tests?
I mean, what’s the error of the error and what’s the sampling error because of it if doctors are feeling this way, It’s either ignorant or truth than it’s unidentified in that that needs to be tightened up.
Or needs to be launched through meaning that A lot of these tests are out of the product.
Or I’m doing a blood test to do, and I need a certain sample size, and I don’t happen to get a blood test to have that but we’re identifying that that thing. I think that might be the one test they’re okay with, but any of the sort of deeper so I don’t know if tell you you can get into this at all or not. I mean, we are pretty explicit that their business plan is to accumulate all this data to guide from two trials.
They’re at the Hiddel. Yeah. So I mean, I think with respect to this, I think drug discovery.
So I believe it’s Roche, for example.
You know, they’re they’re really trying to pioneer this.
They’re looking to this bioinformatics approach.
How do we get the proper medicine to the proper patient within the proper amount of time?
And so they may be looking to, you know, acquire this information for their purposes.
And one thing I don’t think we brought up was insurance companies.
So, you know, just health insurance in general.
Is that fair.
If we know through your health record, Carter, through the data that you have submitted on a in the future, a standardized platform that you are going you or you’re more inclined to develop some type of cancer when you’re over fifty years old.
And we have the scientific data. We have the information to essentially prove it, are we going to charge you more for insurance or not? Well, so this is the flip of business model we’re thinking about.
Well, I go to that insurance company because they found it first and gave me the information and so there’s some value to ride from that.
I mean, it may be a little confusing about how that value is delivered. Right.
But We do know that Novartis and not I Norbert Nordisk has thought about creating a diabetic specific insurance plan.
Where they would specifically target diabetic.
Now I’m not positive how they would pay for it, but it might be that they say, look, We know that if you eat rice or use these behavioral things that your risk is mitigated in. So we will provide the other behavioral incentives independent of clinical.
Right. It gave you in that mindset. Right.
Because we understand behaviors, we can pull a whole cost cycle.
It’s nice. I’m sorry. So they’re they’re who pays for it maybe from it has to be worked out, but that’s a different kind of delivery model. So if somebody’s got a genetic predisposition, can that can we flip this?
Is there anybody fiddling with a business model that changes that from a negative or positive that, oh, you’re a rate being kicked me out to I wanna live and you’re gonna provide me a low cost way to avoid in it. Does that even avoid so we heard in the the other conversation cancer people is if we can improve detection.
We don’t need to do stage four cancer discovering longer.
Because there there will be so few people getting into that stage cancer as well as the detection improvement — Right.
— that that will wipe off.
Right? That’s for I mean, I I know that sounds pretty frugal.
But what if what if somebody could provide the economic solution or prevention? There’s one that got section, but what about Right?
What is the detection of prevention?
And that’s sold as a solution that says, we sequenced you we do these things. You’re environmentally inherent, so we’re gonna sell you a prevention program.
There’s somebody buys and it cost them twenty bucks a month to be on a prevention program.
And so they never turn into a two hundred thousand dollar picture. Right.
And sorry. Go ahead, Craig. I was gonna go in different directions.
So you’re getting I think what’s really interesting is I don’t know if you really understand bioinformatics what it is. I still don’t, to a certain degree, you’re still learning. But it sounds like this may be the process by which we can really achieve that. You know, maybe not in the next year or or two, but ten years, twenty years down the road.
You can take two groups of people, healthy people, and people that have a certain type of cancer.
And then through biom informatics through statistical analysis.
You can identify, hey, do these people have a certain genetic trait in the healthy group where they may have a a trait for this cancer, but it is dormant. And these people who have this cancer, are are are experiencing experiencing and and showing signs and symptoms of this cancer, where can we find you know, a trend and where can we identify why this person has the trait but didn’t get it and why this person has a trait and then did get it.
I don’t know if that makes sense. There’s a pretty good But There’s a good head talk on it.
Relative to the whole topic.
And if anybody is on the line and wants to contribute, please feel free.
We’re obviously more less in the presentation mode and more in the kind of discussion mode. But from our perspective, I guess the question becomes, we’ve seen a number of different businesses in this space, not necessarily from a multiomic perspective, but we invested in cofactor on the basis of personalized medicine and their ability to hopefully identify better patterns from the deeper RNA data that they have in the dataset to help with an oncology standpoint.
You know, and the I guess from my perspective, you know, making it actionable for us, do we think there’s a bigger play from a, you know, from a a population health standpoint or better plays from an individual disease state standpoint.
So for example, again, if you go back to cofactor, It’s RNA analysis on for oncology right now, you know, flywheel, is all neurological at this point in time.
They aren’t using that platform as a data discovery platform form, but they have the data to be able to do it.
Right now, it’s just a collaboration platform. Or yeah. It’s all neurological. Installed image based neurological at this point in time.
Or – but then you go back broader tension metrics was taking data, hospital level data, and looking at aggregating it to identify patterns for larger ways to guide treatment, right? So it wasn’t that they weren’t predicting that you had you know, X, they were telling you when you got X, you know, this is the way you should go get treated. Because these, you know, across the population, we’ve seen better outcomes with and at a lower cost from people doing getting this kind of treatment.
And I guess so from my perspective, I’m trying to figure out where, you know, where is there a chance for us to win? I still believe that cofactor is a good investment.
I do believe that Flywheel needs to think about you know, looking at how they take the data that they’re getting and add more of their own analytical capabilities to it.
I think the same thing with gene matters.
I think gene matters is in a unique position from a data perspective to have to perhaps get and deliver insights that others don’t have access to.
You know, Teddy had pattern up there, which is a company that popped up a week or two ago through some stuff he was doing and I had read about they’re in the pattern looking at massive unstructured data sets and trying to deliver to identify patterns in the data that hadn’t previously been seen and draw linkages to them. So you know, certainly as we talk about these datasets, having someone who, you know, If we believe that in and of itself is a unique skill, right, that is gonna be difficult, pattern identification is something that’s gonna be difficult for people to do inside these these data structures or unstructured data should we be looking at them as, you know, potential enabling technology for, you know, this kind of for for this kind of analysis.
So where, you know, it’s a question of Will there be a one off provider or will Google just keep processing it so we can should just let them do that as they generate the answers, invest in the the venn diagram of their answers, and they Wonderful.
So yes. So there’s you’ve got some combination of somebody somebody’s gotta aggregate the datasets.
Right? And that’s gotta be aggregated cost effectively.
To Mike’s point, some of the org dot orgs are gonna be able to do that cost effectively.
In health care, I look at our investment in Aggrable. Aggrable we we invested in Aggrable because the CPGs were paying for Agurable to essentially aggregate all the data from the farmers.
So there was no cost to the farmer to throw their data in And then we’re gonna have a call.
We’re gonna make a tonality because, you know, we’re gonna have ten things.
Now we’re gonna have fifteen. Now we’re gonna and they were getting information from the MPG.
So from a dot com standpoint, that’s the model. So how do we I mean, who Is there a an equivalent in healthcare of CPGs paying, you know, aggregate to aggregate this data, which made it effective for the Well, let’s say, are the drug companies paying for population data yet?
There may be more than that, but You mean paying for it from from vendors who offer that?
Or Yeah. We’re just acquiring it.
Maybe both or one or the other.
There’s little to no cost Sorry. Go ahead. It’s a it’s a good question.
I mean, I mean, I’m a data aggregator too. Well, my data that I aggregate is, like, quote, unquote, free.
In other words, the source doesn’t charge me for it, because I just do public information.
But for, there’s there’s also a lot of four sale databases out there, especially if you when you look for what the drug companies are interested in for purchasing and all of that. That’s that’s really been That’s really been what they’ve done for years.
So I’m I don’t know. I don’t know who the players would be, you know, particularly in that data for sale, health data for sale unless you’re talking about the insurance providers, because they do a little bit of that.
Interesting. Well, it feels like that. The agrable.
And so in our internal language, who’s the agrable?
We thought we looked who was a company out there in Cisco. We were looking at the the the son of the guy had gone to school with David, It was the consumer play – consumer genetics play where he was getting people to give them genetic data and population and their, you know, their regular health data and their blood work.
In exchange for customized vitamins.
Oh, that guy that we but it was a little bit black magic.
Yeah. I can’t remember. But that was you know, they were he was getting you know, he was getting consumers to pay him to take their data.
I mean Well, this is a model that’s worked elsewhere in other businesses.
I mean, this is a little bit of the conversation I had when we did the Andrew Well meeting.
And and and David Christianson, have you been tracking when Andrew Lowe’s been proposing with the megabund?
No. I’m not terribly familiar with it. You remember Andrew from MIT? Yeah.
Yep. Yeah. I just yeah. So we should talk and I should update you on what he he was in town, like, a week ago.
Oddly, I didn’t get invited by Watchview. I got invited by MIT. It was really awesome conference.
I should have invited you. I’m sorry. Or you should have been invited by MIT.
As we look at these patient data, there’s no HIPAA rules in China.
And so China is aggregating the data footprint. Is that right? That’s UAE is is sequencing every single citizen DNA at birth to to sort of set a baseline and may get a public good.
Who are we talking about? Was it the Norwegian company that we’re — is in Norway or the Netherlands also had very progressive health policies around data? Because there was that – I mean, if you want to write the drop very, very, very liberal rights right to try schools. So phase one drugs can be applied to a stage four cancer at at the patient’s request.
Okay. Sorry. To business model, Yeah. So I think the Agurable we in our in our parlance, the Agurable business model is interesting in this context from a dot com standpoint. Mhmm. Right.
And will Google provide all the data just make it available. And some business model, you have to be determined that they can do what they do right now.
So widely available, easily accessible, lots and lots people can do it, and somehow the economics work out in probably some Silicon Valley and miracle.
And if so, if it just sort of aggregates into them and the monopoly, does the data then become available for creating the because people create systems and send them around Google all the time. Right. So what will those And then what do we do about things like China? China is gonna make all the data massively available without a HIPAA restriction.
Well, I mean, the predicate there that’s resolved, you know, we’re getting the data normalized with no hipaa restrictions.
You know, I don’t know if that’s ten percent of the problem or a hundred percent of the problem. It feels like a hundred percent of the problem. Does that then mean we should go do go there to do patient process.
Or will that, like, five years from now when we go like, hey, I go there. Or will Gen Zee show up and be quick to I don’t care about HIPAA. I gotta like I’m gonna put my sequence to PDNA on the on my Twitter feed.
You know, do whatever you want with it. Right. So the other two thing two things that Benson Hill and Agerville have both done very well to improve the models is consume and normalize publicly available data — disparate data sets.
I don’t know, again, whether or not there’s somebody I have not seen somebody out there doing that.
In the healthcare space. I don’t know whether or not that’s a data set that somehow cofactor could start to do themselves or where that there’s I mean, there’s gotta be between everything that’s been published. There’s gotta be somebody sucking in, all that stuff. And you know, to try to to make it available. But There will be, I think, if you look at big data applications.
You need a an intermediate domain specific filter. So It’s logical to use a neural net to understand variant, covariance, and outwires.
But then to interpret the outlier correlation, correlation causation that you need to get back to the physics or the biology could say, you gotta put a domain specific analyst in the middle of that, just sort of say, because big data can’t deal with that. I mean, if you’re doing a wide flaw, there’s still too much noise in the covariance. So you would need some big domain stuff like that. So we understand farming. We’ve got all the data, and we can understand it within the boundaries of some known constraints, which is why I spy isn’t So a program on the phone, I spy, I’ll say it again, is a clinician focused on press answer who basically went out and got twelve centers to sign up with her, and she she has a single a single form, patient subscription system. So the drug companies can come to and say, I want to go to the twelve centers, subscribe, two thousand patients into a breast cancer trial.
And she she has all their historical data and merges that all together in a particular way. And all the all the principal drug companies focused on breast cancer, signed up with her except Dray and Dray.
Explicitly refuses to do business with us.
But what she’s starting to do is she’s now starting to bring in diagnostics and other things around her, and she started to look at startup to figure out how she can use them in the platform and she’s becoming translational clinical test bed for research that would be domain specific. So if you then sort of attach the data stream to her, And after that, what she’s doing fits my model of how you translate technology based on just my experience.
And it’s an intriguing, you know, those are the path places we should watch for to go find the company. Is that the the test bed place to find the companies? Why would you why couldn’t we just put one together ourselves?
Well, there’s only eleven of us.
But, I mean, I I think that she as much wants to externalize that kind of approach. And maybe on a leverage rate. So there’s opportunity to probably partner in some fashion.
Because the other constraint we have here is is that the sand researchers are so purposely violent.
That is very difficult to figure out the opportunity if they’re not in share mode.
Does anybody know why Researchers are so reluctant to share compared to other industries.
I I think they further increase share more broadly and make more money.
Okay. That’d keep going or we’re over time. Yeah. We’re done.
So the last slide is just are we on the last slide?
This is the analysis.
Very, very high scale.
So should we invest four conditions? We need to really understand the synergies between cancer and multiomics bioinformatics.
Don’t be signed up, don’t be too inclusive of cancer. There are far more diseases that this can be utilized for. Meet advisors and subject matter experts identify strong lead investors in this nascent space.
And like bottom line, we have to understand that this is a computer science play.
And Yes, I’m not sure. I agree with that picture going.
Opportunity It could be right. I just tell you you gotta convince me. But Okay.
So obviously, this space is to consider precision medicine, bioinformatics, translational bioinformatics.
There’s companies out there that we just do not know about that are actively pursuing this because they’re in stealth mode. So one thing on that is what message do we need to transmit in our marketing to get the self help companies to come to us?
Do you want my honest opinion on that?
I prefer to be honest.
Okay. So marketing in hand, we really need to go out and start tapping some of these early thought leaders.
And bring them into I Select and how do we really brand ourselves an expert in in this type of field. I mean, our conversation won’t get into it here that we had yesterday.
You know, bring in the doctor from Stanford. Bring in the the the doctor from Michigan from Northwestern from from Harvard and MIT, bring them together and build something, you know, meaningful. And why why does somebody else have to do it? Why can’t we do it? Why can’t it’s maybe stupid. Why can’t we put together a free podcast that that talked about this.
Put together Well, so on that front — I think — — even with I don’t know if David and Luz are on the call, but I think we need to freeze David what we saw with Andrew’s presentation last week. Right. They wake and collect them into those forecast. Right.
Because I mean, but why are they going to, andreessen horowitz? Why are they going to some of these massive VCs while they’re in stealth mode and secretively pitching them what they’re doing, and there’s no sign of of of, you know, funding.
I know of a company that I, unfortunately, for privacy, I can’t name because they haven’t come out in public yet, that is doing and they are actively pursuing venture funding.
Great. Need to be known. People need to come to us. We need to to let them know.
Okay. So we did not talk about this.
Aside from one company. One of these ones say is the same pattern applied. So all the other to these classes as well as buy them.
Say biome’s got a similar kind of nebulous sort of science based translational base.
That and there’s an overlap between biomed and cancer. So it’s not just separate the nicest of the cancel time. Of course not. Yeah. Definitely. And which may mean that there may be some other pathways that are more easy.
What you were saying? Go ahead.
So we looked at one company called January.
Which is more of a b to c opportunity.
Really exploring a platform for personalized medicine.
It’s application based on iOS platform. I do not know if it’s been released yet. But anyways, So I Craig brought this up, companies to consider. I’ve got a call with pattern computer tomorrow, incredibly interested in speaking with them.
And just to kind of jump in kind of throw on top of what Craig said?
The big thing here is it’s this is not pattern recognition. We don’t train these computers to go out and and look we can go out and look for specific things.
Because at the end of the day, we don’t know what it is that we’re looking for. And what pattern computer is doing, and they’ve got a a study, I believe it’s attached to their website that shows this. Again, Saturn computer was operating in stealth mode until, I think, March or May twenty second or twenty third.
Anyway, so pattern identification, pattern discovery is the big thing. How do you find something in a giant field that you don’t know what it is you’re looking for?
So that’s really kind of the approach that they’re taking and very interested. So anybody who’s currently on the line, if you know of some subject matter experts or potential advisors within this field and you wouldn’t mind sharing them with us.
Would really love to learn more and dive deeper into what some of these companies are doing.
Because in my analysis, again, my personal opinion, as far as precision medicine, this seems to be the pathway. So I’m gonna give you two test conditions to think about — Okay. — by h pylori I’m sorry? H pylori.
You can you can figure out what that is. But why didn’t we realize he was right? Twenty five years ago or actually thirty five years ago. Okay.
The other is we identified The first known observance of immunotherapy was in eighteen ninety.
So a doctor observed and documented. The cancer was eliminated when somebody got influenza.
And noted it.
And it’s from eighteen ninety to about two thousand five. That’s how long it took for us.
To recognize what was going on. Further to that, it was really aids Ayd made a shift in our discovery process for genetics.
That at the time was statled and was just by the researchers.
And so there’s a series of transitions that went about in terms of how people observe things.
Advice to the point that they realize being able to do genetic sequencing, using that as a base and using that to identify solutions resolving in the immunotherapy that it’s a transition. And it took us ‘nineteen, ‘eighteen ninety to two thousand and five to make clinically relevant. Right. So is there any can we learn from that lesson to understand between bioinformatics clinical behavior, rules of procedures, business models as to what helps consolidate that time.
Okay. So use those as you’re thinking through this research.
Those are two examples test conditions to go back. Would say, is this have any impact on those two conditions.
H pylori and and and mean a fair Okay. Anybody on the call have anything to add or questions or insights? Or Alright? Thank you for your time. Thank you very much.
Sure. Thanks. Appreciate it.