SYNAPSE: Tell us how Ymir Genomics was founded.

PENDERGRAST: Ymir has an atypical story. I had been working at Archemix Corp. in Kendall Square for ten years until the company failed for business reasons, despite its excellent science. Around that same time, my brothers, both serial entrepreneurs, sold their e-publishing company to Barnes & Noble. Given that all three of us were looking for new challenges, we joined forces to start a biotech company.

This is an atypical start because: a) biotech companies aren’t usually a family business, and b) we started the company without knowing what we were going to work on. No business plan, no widget to sell, no patents. Thus, I spent the first six months just reading papers looking for unmet needs and deciding what we were going to do.

PENDERGRAST: Since I was an RNA biologist, I read up on the latest in the RNA field. I was interested in microRNAs despite the fact that I had no experience in that sub-field; I suggested doing some bioinformatics in that area since my brothers are information scientists and could become involved. With the help of my good friends Mark Histed and Fred Huettig, two MIT alumni, we wrote a series of programs that sifted through deep-sequencing data towards discovering novel miRNAs linked to disease. As a proof-of-concept, we applied one program to a public database and asked, “Can we find things with our program that other programs missed?” And indeed we did! We found several microRNAs that we are now writing up for publication. We also found some microRNAs that could be linked to disease so we brought in another friend of mine, PhD biochemist Anna Markowska, to generate experimental data showing that they were upregulated in liver disease.

Our next step was to apply our program to a Next-Generation Sequencing database created by our own hand. Our lawyers told us novel RNAs discovered from our own database, rather than a public database, would be easier to protect. So it was back to the literature! After some reading we came up with the lowest hanging fruit that I could see for diagnostics: bladder cancer. Initial treatment of bladder cancer can be simpler than for other cancers; you generally just cut it out. The problem is that there’s a high chance of aggressive relapse. Survivors of bladder cancer have to get an invasive cystoscopy twice a year for the rest of their lives. So there is an obvious need for a less invasive diagnostic to reduce the number of cystoscopies.

Now the problem with low hanging fruit is everyone will go for it. We decided we need something else to distinguish us from the many researchers attacking this problem. The answer was extracellular vesicles or exosomes. Tumors pump out millions of extracellular vesicles a day, and those vesicles contain biomarkers. For bladder cancer those vesicles are going straight into the urine.

I was incredibly naive and thought, 'Oh, this should be easy – acquire some bladder cancer urine samples, isolate the extracellular vesicles, purify the RNA, perform deep sequencing, then run our program to look for differences in microRNAs (and novel microRNAs), and we’ll get a great biomarker.’ We spent $30,000 buying bladder cancer samples and then tried to isolate the extracellular vesicles. We couldn’t do it; our yield was too low. We then bought several commercial kits to isolate exosomes from urine, but still couldn’t do it. Later on we discovered that everyone in the field knows that most of those kits don’t work well for urine.

At the same time several Supreme Court decisions, including the Myriad case made patenting genes (potentially new microRNAs) virtually impossible. So our original business plan of patenting newly discovered diagnostically relevant genes became very challenging.

People always talk about how useful it is to switch fields so you can bring a “new perspective” to it and I agree with this concept. However, they seldom tell you to be prepared for some serious growing pains. So this was our “growing pains” low point. Our original business plan was not viable and we were stuck with $30,000 of seemingly useless bladder cancer samples.

PENDERGRAST: It’s almost a cliché that a little company can be nimble. We dropped the bioinformatics and instead spent another year and a half perfecting all our methods for urine samples - we’ve since expanded into blood and saliva. Currently, we are in the early stages of negotiation with companies interested in partnering with us to license these methods and make them available to everyone - including the recently interested microfluidics community! Once we generate funds from the license, we can use it as funding to find biomarkers. Even though any such genes can’t be patented, one can still build a company off of biomarker panels and the methods to assay them.

We are working on publishing an article showing that urine is especially easy to work with for microRNA biomarkers, but at the same time we are expanding into other biofluids. Given that the market for urine diagnostics research is very small, we are expanding into more popular fields such as blood. We aim to become a biofluid sample prep company rather than just a urine-diagnostics company given that urine research market represents only 3% of the total exosome diagnostic field.

SYNAPSE: How has the unique relationship with your business-oriented brothers influenced Ymir’s development?

PENDERGRAST: Even though I have been in industry for ten years, I am more of an academic at heart. I was at Cold Spring Harbor for seven years doing two postdocs, so my tendency is to never give up on a problem. Being more entrepreneurial, my brothers are of the thinking that, "It’s not giving up. It’s just switching to something better!" My brothers advised that our business model wasn’t a good one so we came up with a new one. No harm, no foul. That type of thinking is a lot different than the academic mentality of pursuing something relentlessly until it works. I think it was easier for them to convince me to switch because they are my brothers and I trust them implicitly.

SYNAPSE: Last year Ymir announced a collaboration with the Medical University of South Carolina (MUSC). How did that come about? And more generally, how do collaborations in industry arise?

PENDERGRAST: It came about from a connection made at a conference. The man sitting next to me gave a talk right before mine and described his frustration with isolating extracellular vesicles from urine and how none of the kits worked. I was sitting there thinking, ‘Wow, he just handed me the greatest segue ever!’.  After his talk, I started my presentation with, "I’m going to solve your problem Dr. Drake - and we’re going to do business together!" It turned out that he was in charge of a biorepository for urine - talk about a match made in heaven!

PENDERGRAST: There have been some changes that have made it much easier to start a company without much money. The idea of biotech incubators like LabCentral and Cambridge Biolabs is really clever – they make it almost as easy to start a biotech company as it is to start an app company. The biotech world has been jealous of the computer world for so many years because of the clichés like, ‘Facebook was started in my dorm room’, ‘Microsoft was started in my garage.’ But you just can’t do that with biotech. You need a lot of equipment, experiments are much harder to run, and it takes a significant amount of time. The time from when you start your company with an idea to when you actually get the payoff is longer for biotech than for tech companies. Also, the costs of running one contract research organization (CRO) study with animals is so much higher than buying a server and one programmer. There’s no comparison.

It used to take $300,000-400,000 to really start up a lab, and now at LabCentral, it’s $4000 per person per month - you get a bench and it has pretty much everything you need. You have to buy consumables, but that’s just another couple thousand. The availability of LabCentral and other shared lab spaces is one huge component that has allowed companies like mine to get off the ground.

Another reason why it’s much easier now than in the past to start a company with almost nothing is because of CROs. You don’t have to hire a full time employee to work on certain tasks - you can just hire a company. If you want to do a proof-of-concept study, but you don’t want to hire someone for $150,000 a year to do it, you can instead spend $40,000 for a one-shot deal and get your proof-of-concept from the CRO. Websites like Science Exchange can help you find CROs to do almost anything you may need.

In addition to places like LabCentral and CROs, the third reason why it has become easier to start a biotech company is that right now the money is there. Look around Kendall Square - there are buildings going up all over the place. VCs are always looking - they see 10 companies a day and don’t like most of them. Big pharma is insatiable - their pipelines need constant replenishment.

PENDERGRAST: It used to be that big pharma wouldn’t find out about a hot new company until they read about them in BioWorld. But now, big pharma must be much more proactive. The competition is fierce. One way to find companies of interest to them is by sponsoring incubators like LabCentral or even by running their own incubators. This way they can gain access to startup biotech companies much earlier. For instance, a sponsor to LabCentral will run a ‘Meet and Greet’ event to learn about the resident companies. That information is very valuable. Starting a relationship early with a startup can even put them in position to influence the direction of a company towards acquisition.

PENDERGRAST: One path for biotech startup companies is to write Small Business Innovation Research (SBIR) grants from the NIH, NSF, and DARPA. There are at least three companies at LabCentral that started with a scientist who had an idea, got a little bit of preliminary data and wrote an SBIR grant to get the $150,000 or so to start a company. Then they rented some space at LabCentral and obtained enough follow-up data to get a Phase II SBIR grant for over a million dollars.

However, this is not an easy path. These SBIR grants are very hard to get; I had extensive experience writing fellowship and grant applications as a post-doc and had some success. Yet, Ymir is 0-for-4 on them. A very high percentage of SBIR grant applications are turned down on the first try, but once you’ve learned what they don’t like about a given proposal and rewrite it, the success rate goes up significantly. My first two attempts were not even scored although the last two were fairly well received and I am hopeful for the resubmit. Another great thing about LabCentral is the cross-pollination among the various executives, employees, and companies. People are very open and I’m able to pick the brains of the incredibly talented CEOs and scientists there. This has been especially valuable for SBIR writing. Not only for writing advice but also simply for morale. I recently talked to someone who had just received an SBIR grant and was heartened to hear her say, "0-4? Shannon, don’t be discouraged. I was 0-for-8 before got this one!"

SYNAPSE: When and how did you know you wanted to go into industry? What should a graduate student consider when deciding whether or not to do a postdoc if they’re fairly sure they want to go into industry?

PENDERGRAST: Before graduate school, I spent a year at a small biotech company and I loved it. We were one of the first to clone the cytokine GM-CSF; doing work there was so much fun! So when I went to grad school at Rutgers, I told my advisor Richard Ebright to give me an industrial project. He had the idea of making artificial restriction enzymes - a very entrepreneurial project. We used a Catabolite-Activating Protein (CAP) from E. coli as there was much known regarding how CAP bound its 22-bp site when in a specific conformation. Based on the crystal structure Richard cleverly hypothesized that if we placed DNA cleavage moieties on the sides of the protein such that they only contacted the DNA when the protein was bound to its specific DNA site, we could generate DNA cleavage at only those sites. That project ended up working very well; we finished around 1990 when the Human Genome Project was just starting and had people looking for ways to cut DNA into larger pieces than permitted by the 8-12mer sites that E. coli enzymes cut at. The work made a splash, and we had a nice Science paper.

Then, I went to Cold Spring Harbor Laboratories to work with Nouria Hernandez. CSHL places an emphasis on academia; industry was looked down upon back then as a career path. I was receiving world class training from Nouria and others but I let myself get sidetracked onto the academic route for a while, and I had some papers from my postdoc but nothing that created a good model system that I could take to an academic job. So finally after 5 years I decided to go back into industry. However, I realized the techniques I had learned during my postdoc were not useful for industry. They were all biochemical transcription-related and industry was not interested in these techniques. Nouria was very gracious to allow me to basically do another post-doc in her lab training with members of the Spector cell biology lab where I learned microscopy. This helped me to land a job at Archemix in 2001.

So to answer your question; sure, do a post-doc, but pick the right one and stay focused. I delayed my career for a few years because I went back and forth about what I wanted to do. I was thinking ahead when I was young in graduate school, but as I got older, I lost my way. I wasn’t really sure what I wanted to do and just coasted, pumping out a minor paper every year, and being happy with that but not really thinking about the big picture: ‘What is my next step?’ and, ‘Is this getting me to my next step?’

I highly recommend you enter a lab and pick a project that enables you to acquire a skill set that industry might want when you’re finished. I imagine it’s much easier to do that at MIT where there are labs with an industrial bent that spin out companies or are working on things that are commercializable. If possible, you should work on projects you think industry will like. How do you know what those might be? Look at the job listings companies put out when they’re filling positions. See what techniques companies want, and match those up with what you want to do. It’s a combination of what you like and what’s in demand.

SYNAPSE: What advice or words of wisdom do you have for graduate students or post-docs who anticipate starting their own company or going into industry?

PENDERGRAST: Again, it’s a lot easier these days - both at MIT and elsewhere - to take advantage of opportunities like joining the MIT Biotech Group, talking to VCs, and getting a tour of LabCentral. But more than that, think of projects that end up applicable to an entrepreneur. Talk to your advisor about it. Some PIs couldn’t care less about entrepreneurship - they’re totally focused on the academic realm. But go to their talks and read their papers, and if you hear of something you think could be an idea for a company, don’t hesitate to approach them and talk to them about it. For example, I have a friend who’s a PI at a major university who had no desire to start a company, but she had a cancer treatment that was really quite marketable. Her post-doc asked her if he could use the IP they generated to start a company and she said, "No problem! Of course, the University and I will need to be included. But I am fine with you taking the lead!"

Ideas are out there, all over the place. So every time you go to a meeting and hear a talk, think, ‘Is there something here that has value, and if so, could it be brought to market sooner if it was spun off into a company?’. There’s a reason why Bob Langer starts all these companies – sometimes the ideas can be turned to practical use much faster if they’re spun into a company where people are completely focused on the product and not just writing a paper. The second question is, ‘Am I the person to start that company?’ That’s usually the harder question, but often you can be that person, especially if you go into a lab and are up-front with the PI right away.

Also, there are a couple of blogs you should read - BioWorld and Matthew Herper’s blog on Forbes. Those help you get a feel for what’s out there, what’s raising capital, what’s successful and what’s not. I didn’t start reading BioWorld until after I was working at a company for several years. Of course your main focus should be your graduate or postdoctoral research, but if you can afford an hour a week to do this type of research, it will pay off in the long run.

And finally be persistent and, most importantly, adaptable!