This post was originally published on this site
Aldo refers to himself as a serial problem target, meaning problems find him and he starts companies to solve them. He has started companies in many different industries including Life Sciences, FinTech, Blockchain and Media. He is currently the CEO and cofounder of Intervenn BioSciences, a company dedicated to diagnosing cancer through glycoproteomics.
Alan: Can you share a little about your background?
Aldo: A lot of people refer to them as serial entrepreneurs I’m actually not that person. I did start a number of companies but I see myself as a serial problem target. Problems keep finding me and I try to make companies to address the said problems. I’m a software engineer by training and I’ve started a number of companies, in blockchain, in FinTech, in digital media, and even in that matchmaking. Essentially I had a of successes in the past and people often ask me how did I get into Life Sciences. Without belaboring a great purpose, it’s also a terrible story. My family was decimated by cancer. I think in the past two and a half almost, three years I’ve been asking a lot of questions around what’s going on with oncology. What’s going on with Diagnostics, what’s going on with therapeutics. The most recent venture that I started is called Intervenn BioSciences. The genesis of it was a lot of frustration from me and from some of my co-founders.
Alan: With no prior experiences in Life Sciences, how did you build a team?
Aldo: I’ve never started a company and then went in the same industry for the past 15-20 years. The only commonality I have is absolutely simple, it’s about the team. It’s how I surround myself with subject matter experts. Some of them are some of the deepest subject matter experts and I also I think that one of the more important things I do is that I understand gaps with our workflows, with what they have what they know and I try to understand who they can work with. For example, we’re interrogating something called the glycoproteom. Your proteins have these accessories and one accessory is a sugar. So this is actually a very content and very information-rich structure. Once interrogated it can actually give you an early diagnostic for a few other diseases not just with cancer. The problem is, to interrogate it you need a machine called a mass spectrometer. These giant machines do characterization of different molecules and compounds. They’re commercially available but there’s one caveat- they generate a tremendous amount of information and traditional by statistics interrogating glycoproteomics sometimes it takes months, even years to do analysis. When I found out, my old co-founder Dr. Carlito Lebrilla of UC Davis was telling me that he was running a cohort of 100-200 patients. He told me that the data analysis alone took between a few monthsup to a year. So I said, this is a ripe for artificial intelligence, so why not create a neural network that basically just speeds up the whole operating process. I’m happy to report from a few months we’re now we’re down to 12 minutes. People often ask me how I get teams to work on this. It’s basically identifying a unified problem and then seeing the different areas within the value chain that need integration. Even if you had the most powerful mass spec which created you know hundreds and thousands, even millions of features, you couldn’t manually create that, you couldn’t curate it and you couldn’t run it on statistical software- it would just take way too long and I’ve seen that process happen. and my co-founders Dr. Lebrilla and Dr. Bertozzi have gone through it over the years. A lot of people ask us, is having artificial intelligence just an excuse to be cool? It’s absolutely not. The reason why we needed a neural network was because you know the data sets we were generating we’re so ripe for vision that there was really nothing else we could use besides a great recurrent you know network to interrogate that information.
Alan: What is the mission of your company?
Aldo: We want to make sure that we create a world where no one is ever blindsided by disease. Everyone seems to get blindsided by cancer, Alzheimer’s, they don’t know even know they had it in their family history. The other one is to provide tools that can intervene when needed. That’s really our purpose. We want to give physicians access to tools -that have actually been lying around but have never before been created as part of a clinical process.
Alan: There was an article recently on China you know creating the next superhuman what are we going with the AI and the technology in this area and is that a real risk?
Aldo: I think for us the next generation we see is human digitization. Before we talk about singularity or human augmentation or you know man and machine I think we need to understand ourselves. Human digitization is this move where you want to omic yourself. If you think about the human spectrum there is something called genomics which is of course our genes and you know all the way to post translational modification. If you think about that entire spectrum, if you think of biology it actually goes through that. Genomics goes to transcriptomics which goes to epigenetics which goes into proteomics which goes into epiproteomics which goes into post translational modification- that is the picture of life. A lot of companies right now talking terms of multi-omics to understand human beings. Think about it as different types of data sets, different types of information about a certain organism or a certain biological sample, but seeing through many different lenses and in different tools. When you understand one, you supplement the information in the other. The gene BRCA1 and BRCA2 are associated with the development of breast and ovarian cancer. Genetic tests like once you can get from 23andme or colored genomics you can tell out if you were a carrier of these genes. If you actually were a BRCA1 one positive you would have a 40 percent likely chance of developing cancer. So that’s very unfortunate but what’s interesting is how you know marry that information with proteomic information. For women who have these BRCA gene, how do you track the development of those genes into cancer? Some people say using glycoproteomics you can actually track very very early stages of tumor development all the way into stage four.
Alan: What area of cancer diagnostic are you focused in on?
Aldo: Our lead asset is a clinical diagnostic test for ovarian cancer. We’re doing clinical trials for it this year. We also have many other assets in the pipeline.
Alan: What’s been your experience as you’ve combined AI and machine learning with Life Science?
Aldo: I find that AI and machine learning there are big hammers. A lot of people coming at it from the AI a machine learning point of view are looking for giant nails sometimes without respecting life and it’s complexities and timing of life sciences. I have to admit when we started this company two and a half years ago, I came into this industry thinking that I could AI and machine learn anything and I could figure out what it was in life sciences. I was quickly proved wrong because there is a certain complexity, a certain way of gathering and interrogating information in biology that needs a lot of rigor outside of machine learning and artificial intelligence. So for us we actually don’t call it AI, it’s actually augmented intelligence. We’ve been saying this for the past two years because what we want to do is augment scientists. They already know what they’re doing they already know what they’re looking for, they just have to comb through hundreds and thousands of [bits of data]. We’re not trying to find the signal or amplify a signal that’s not there, we’re amplifying a very very very strong signal but finding it really really fast. That’s an important distinction of what we do, why we use AI and why it’s important again. It’s not a shiny object, glycoproteomic information is so information rich, it’s groundbreaking- it’s just incredibly hard to interrogate because of the sheer amount of information generated.
Alan: How hard is getting through the regulation in this country versus other countries?
Aldo: This is my second regulated industry. My first one was in FinTech, where we had to register as a money service business and money transfer licenses in the United States, and a few dozen countries outside the United States. The regulatory process is a necessary process- that’s what it is. You embrace it from day one. Just like FinTech, just like monetary policy health, has its own caveats, there’s a little bit more strictness to it because there is personal information, there’s a lot of information relating to individuals, there’s a lot of information that are proprietary to disease and institutions so there are silos of information too.Like in my previous company, we embraced compliance and regulation from day one. We have the same type of mindset here. In China, they’re very very strict with health information especially sharing information. In the United States, you you follow multiple pathways, you follow multiple regulatory oversight agencies, to register a clinical trial. To register your product you go through a certain set of parameters when you do your validation study.
Alan: Currently where is your company?
Aldo: We’re going through global clinical trials. During this year, 2019, we’re trying to do a validation trial for our ovarian cancer assay, so we’re actually enrolling around 2,000 women from three countries, primarily from the United States but also from Malaysia and the Philippines. Our preliminary information the data that we generated may be a year and a half ago outperformed every major test for ovarian cancer and it’s actually not magic, all we did was commercialize the work of Dr. Librilla and Dr. Bertozzi. Again we supercharged it using next-generation mass spec and AI. They’ve been doing this for a very very long time, it just took them a very very long time, that’s how we were able to really really bypass the whole research and development and the whole discovery process. We have multiple collaborations with the best cancer centers in the country, I unfortunately cannot name them, you can find them on our website. We just raised our first institutional round with some with some of the best investors, I can mention them: Genoa Ventures, True Ventures, Amplify Prado SV, and some we actually had a a wonderful around I honestly wish we could accommodate everybody, but we were very excited with the warm reception and just this understanding of what glycoproteomics can do in mass spec and AI. We’re very excited that we have all of this momentum and but at the end of the day we care about patients. We’re seeing that you know we’re on track to go into clinic by middle to end of next year, that’s great especially for ovarian cancer where there is no diagnostic.
Alan: Where can a person find more information about your company?