In this moderated talk, Sarah introduces her former postdoc José Leal, who recounts his journey from academic research in comparative genomics to multiple biotech ventures. After leaving the Taichman Lab in 2006, José founded a research group in Portugal, led the national bioinformatics core, and helped establish ELIXIR Portugal. He describes his first startup, Comparo, which leveraged comparative genomics to identify non‑redundant bacterial genes as potential drug targets, aiming at veterinary and biodefense markets. The project stalled when efficacy did not scale to larger mammals. José then co‑founded Ophiomics, a precision‑oncology service that built a small laboratory within a large pathology group, and created the OffidiX software to turn sequencing data into clinical reports, delivering Portugal’s first commercial liquid‑biopsy report in 2016. After selling Ophiomics, he pivoted to a new venture focused on diagnostic products for liver‑cancer patients awaiting transplantation. Using ensembles of support‑vector‑machine models on gene‑expression data, the company developed a kit‑based decision‑support test, secured an EIC Accelerator grant, and launched a prospective clinical trial recruiting patients. Throughout, José reflects on the challenges of fundraising, the shift from service‑based to product‑based business models, and his evolving role as “chief everything officer.” The talk ends with a light‑hearted exchange about his continual need to raise capital and his preference for product over service delivery.
Chapters
Okay, next talk we're going back really far in time to my first ever postdoc, José Leal. And José, I'm not sure if you met Joe, but yeah, José kind of started the early protein interaction work in terms of networks of interactions and American interactions, and then returned to Portugal, start up his own group. And he's also going to tell us about his startup activities. So that's exciting and different, and we're going to have some non-academic— or we're going to have some talks that link to biotech startups, industry, and so on. So this is the first one, and we really look forward to it with excitement. Over to you, Jose.
Okay, hi Sarah and hi everybody. It's— I'm really sad I cannot be with you. I'm sure there's going to be a lot of fun over today. I am not giving a scientific talk. Let me just share my screen. Can you see my screen well?
Yeah, we can see it nicely.
Okay, you can immediately see the corporate look as we keep—
Yeah, absolutely. Yeah, yeah, very professional.
Yeah, anyway, I came to— I left Taichman Lab in 2006, coming to Portugal, and I set up my own research group. And at the same time, I started— I was managing the bioinformatics core and helped set up ELIXIR. And I was the first Portuguese head of node, as these bioinformatics— and I established, or to help the Portuguese bioinformatics infrastructure that I was literally until yesterday, I was the president of the managing board and I resigned yesterday. And I thought of picking up one of these stories, one of the scientific stories that were within this type of activity, but then I thought that probably would be far more interesting for you, at least entertaining, see a little bit of something that I started doing in 2009-10, and that eventually led me to close my group and work full-time in my own company. So I've done a few forays into setting up companies. I'm going to tell you about two, in which one of them did a pivot model, and that's what I'm doing now. So I will discuss a little bit of science, but you'll see it's going to be very, very shallow because somehow I think that my own path is the interesting part of here, but maybe Maybe I'm being too vain. Anyway, so in 2009, we were in the lab working at adaptation to predictable environments that happens for intracellular pathogens and endosymbionts. So, and we're looking at doing comparative genomics, trying to understand the dynamics of this adaptation. We knew that bacteria would lose a lot of genes, but we were trying to figure out which genes were lost. In the end, we published a paper that, while not very cited, it's one of the papers I like the most of my full career, where we show that in fact what you're losing over the course of evolution when you adapt to a particular environment is redundancy, is having additional proteins that can do vaguely roughly the same that the one that you have. So basically you don't have spare tires, that's the simplest image. And we published a paper, went to conferences, and all very abstract. But at some point I thought, you know, these points in which there's no spare tires, these genomic components of which there's no redundancy, are also points that would be very fragile in these bacteria if I was to target them. So one night I went home and literally with a glass of wine and laptop, and that's the wonderful thing about bioinformatics, I started asking, are there drugs that can target these potential fragile points in intracellular parasites. And came up with a very interesting link of drugs that we already knew the mechanism, and decided to try to convince investors to give me money to develop this. So what is basically simple comparative genomics became a tool called Comparo. This is start becoming the salesperson here. But I joined a friend who started understanding what markets would be out there, and we focused on veterinary markets and biodefense markets. These were the organisms that were coming up on our search. We started doing proof of principle experiments, and everything was going really well because every time we added the compounds that we thought were antibiotics to our bacteria, they immediately died, and we were really excited about this. We started developing a very sophisticated IP strategy of which over Hopefully one day over a glass of wine I can tell you all the adventures that we had with embassies and secret services and all that, because at some point we're talking with biodefense domain. And came up with a strategy that we're going to collaborate with the British military, the British defense labs, to do a proof of principle study with one particular antibiotic against malaria, which is something that worries the Americans very much. And then we would sell the science to the Americans and it will be massively rich. That was the idea. So we started the collaboration, and then it failed because the point is when you add it to larger animals, so intermediate-sized animals is fine, but when you add it to bigger mammals, then it's indistinguishable from water in terms of effectiveness. So the issue of effective concentration size never worked out. So we ended up dropping that. We published a paper And I can tell you what we're doing was repositioning a therapy for alcohol addiction into that we discovered was an antibiotic, and we're trying to develop it for biodefense. Anyway, so at this point, there's lesson learned. We took a lot of time to do this because I was never fully committing. I was still running a group and a facility. But at the time, I, for reasons that to this day still elude me, I was, I involved in a very large cancer genomics project. And I had recruited this lady here who is now becoming my business partner ever since. And we are doing a lot of cancer genomics. And one day we realized that there's all this precision medicine fancy ideas coming up and that all the clinicians that we're working with kept thinking this was in the future. And we kept telling them, no, no, this is now. And then they start asking, but where can we do these tests? Where can we study our patients this way? And there was no available company that doing these laboratories. So decided not very wisely, why let's do one. And so we set up Ophiomics version 1. And what we dreamt about at the time was to develop a precision oncology portfolio that would allow us to ask, am I likely to have cancer? Do I already have cancer? How do I treat it? So I have cancer, am I treated, am I responding well to treatment, and so on. And that's what we did over about 5 years. We were very, very lucky because we managed to convince what was the largest clinical pathology group in Portugal— it's these guys here, the Germano Sousa— to let us build our little company inside their big laboratories and use a lot of their infrastructure, including commercial one. And so we focused on setting up what was basically a very small a laboratory dedicated to digital and quantitative PCR, NGS, and so on. We set it up. It turns out that at the time, there was really no software that was professional enough that would allow us to take sequencing data and transform it into a clinical report. I tried very hard to buy one and I couldn't because there was a lot of it was— a lot has changed in the past 6 years, but 6 years ago, you— it was very hard to find a lot of these things that are now commodities. So we ended up investing a huge amount of time in developing software called OffidiX that was used to analyze all sorts of— you can see there MLPA, you can see some Sanger, you can see NGS. And we're using this software to integrate all that into clinical reports. And that's the point that the monster, the software, kept growing. And we already had the limbs attached to it that we're managing all the And just to tell you that in 2016, we produced our first liquid biopsy report. It was the first in Portugal, commercially speaking. I'm very proud of this. Unfortunately, this lady died shortly after. And we ran this pipeline for a few years successfully, I would say, because we managed to reach breakeven. We became very well known, kept being invited to go to national medical meetings. The point is that my life became very much— I was half of my time a salesperson, the other half I was answering the phone to clinicians, advising them of which tests would be relevant to patient A versus patient B. And that was not really something I wanted to do with my life. So in 2009, we said we sold the company, well, actually all what it contained, and decided to pursue something else. And there I'm going to show you a little bit of science again. That started by coincidence, which was we were approached a few years ago by two liver transplant surgeons who came to us and said, listen, we are really good transplant surgeons. We are one of the largest centers in Europe. We are real— we've invented techniques that have changed the field, blah blah. But because we put livers in the wrong patients, they die. So can you help us select the right patients for the transplants? So the specific problem they're trying to solve is if you are a patient with hepatocellular carcinoma, if you can have surgical procedures, they might have curative intent, particularly liver transplantation. If not, and you go to pharmacological treatment, palliative care, it's really bad news. And so the idea of finding which cancer patients would benefit from a liver transplantation was the goal. We looked into these for a while and it clearly was a big problem. Ugo ended up doing a PhD under my supervision in this topic. And what happened in 2019 is that we pivoted the business model of the company. So sold everything. So all the contracts, all the software, all the clients, everything. And we started from scratch. And started playing with research that me and Joanna had developed where we discovered biomarkers, invented algorithms. At this point, we're using deep learning. We had conceptualized the product, but we were a bit stuck. This was mid-2018, late 2018, because we now needed a few million to make this idea a reality. And so I actually went home and lived out of my savings for a year and something, trying to raise capital for this project. We're very lucky that we got one of the first generation of DIC Council accelerated grants. I don't know if you know what they are, but try to think ERC but for startups. And that allowed us to basically start from scratch, build a new company in a new location, build new laboratories. And start developing, to be quick, a portfolio of products focused on the different decisions in the liver cancer patient journey. So I'm just going to briefly mention this product, which is one about liver transplantation. It has a lot of interesting science. So from a computational perspective, this turns out to be— we dropped the neural networks and started working with support vector machines. But this is a series of ensembles of support vector machines that work on gene expression data measured directly from a liver biopsy, integrates with clinical variables, and is able to produce the report that you can see here. In terms of the business model, this is actually, actually sell diagnostic kits to real cell from January, I hope. Obviously, for this to be a success, we need massive amounts of clinical evidence. I cannot just go by mentioning all the other projects. I don't have time to go over all the evidence we developed. I just want to tell you something that I'm really very proud is that we've got a clinical trial approved, a prospective clinical trial that is now recruiting patients, recruiting 8 patients so far. Some of them are patients that will never be eligible for transplantations and are now in waiting lists for, to receive an organ as soon as one becomes available. So this is my last slide, and I just wanted to contrast what my life is from when I was a scientist and I was always raising funds to have the lab going. And now as a chief everything officer of my third company, I still keep spending most of my time raising capital. And now again, after spending the first couple of millions that the EIC accelerator gave us, I'm already begging again for a few more million. So that's it for today. Thank you for your time, and I hope you found it vaguely interesting.
You seem successful at begging and selling, should I say?
Yeah, it has worked so far. Sometimes.
But this one is again, you said never again diagnostics, but then you went back to diagnostics.
No, no, no, never again services. So I don't sell and I don't sell analysis. What I do is I'm making products that I then sell to distributors and then, and then I provide the algorithmic answer to the clients, but I don't have to deal with individual clients. And that was what I was doing.
I see.
I was actually meeting the patients coming to me. And that was something that the previous company did that I was not expecting. We ended up specializing in the desperation niche. People that were already beyond the guideline of transplantation, and they would come to us to ask for hope. And then you have kids coming to ask for hope, and that's very heavy. So that's why I said no diagnostic service, never again. But diagnostic products, in fact, I'm trying to organize a national sort of industrial value chain for diagnostics because it's cool, and there's a lot of bioinformatics and data science in diagnostics, and I find it really exciting.
No, there's a huge amount we can add. Awesome. Congratulations, José, on your amazing successes.
