Monday, 12 May 2003

Biotech revolution

Speaker: Biotech revolution
Keith Williams, CEO of Proteome Systems, was the guest speaker at this AFR BOSS Club event.

Transcript

Monday, May 12, 2003

Keith Williams is a scientist and the fouder and CEO of PROTEOME SYSTEMS, a biotech company based in North Ryde, NSW.

The big bucks in biotech are in the human health area. Agriculture is significant, but the really heavy lifting is in pharmaceuticals. In that space, many of the drugs that the big pharmaceutical companies have in the marketplace are about to go off patent.

So you have a very unusual situation at this time. It hasn’t been like this for at least 20 years, where a group of very big pharmaceutical companies are very concerned about what they’re going to do, because they’re about to lose their competitive edge that has been a licence to print money with some very big drugs. And so we’re looking at a moment in time where people in very large corporations are trying to understand how to change and change quickly.

Firstly, we have a crisis in the pharmaceutical industry. Secondly, we’ve become much more knowledgeable about biology than we were say five years ago. Even today, when you visit your local GP, he or she is working off the knowledge base that involves knowing about one in every fifth part in the body. Think about that when you go visit your garage to get your car fixed, that the mechanic can identify only every fifth piece of your vehicle, and you have an understanding why medicine is a little bit hit and miss at the moment.

In the last few years our knowledge base has changed dramatically, so we’re now getting to be knowledgeable about all of the parts, although it has not translated to the GP yet. The story that I want to tell tonight addresses this new knowledge bank, what it means for commerce and where proteomics - our business - fits in the overall space.

I will start with information and tell you the story of genomics. The information specifying organisms (genomics) involves a four letter code - and there are about three billion letters strung together in the human genome. They are partitioned over 23 different physical chromosomes. We now know that there are somewhere between 30,000 and 50,000 genes or packages of information that specify proteins, the workhorse molecules of the cell.

More than ten years ago it became clear that the way this information would be deciphered would be at the DNA level rather than at the product (protein) level. So the scientific world went fishing into information rather than fishing into what is there (proteins), if you can understand the difference.

For a long time academic groups set out to solve this problem by breaking the information down into component parts and analysing the bits.

It was incredibly slow and to some extent people were content to do it that way as everyone saw this as a huge mountain to climb. Then along came a guy called Craig Venter, who was the “mad genius” rule breaker. He said something like “This is really boring and I don’t want to do it the slow way”. So he made a pact with a company that made tools for sequencing DNA, a company called Applied BioSystems along the lines of “ If you give me a very large amount of money I’m going to solve this problem and I’ll do it in three years”.

Chaos broke out, as you could imagine. All of the vested interests were suddenly being told that they were going to be put out of business and not in ten years but in three years. In actual fact Venter broke the back of the problem in two years

He built one of the biggest computing facilities in the world and he spent a few hundred million dollars. It’s not completely solved, as there are bits and pieces not joined up, but to a large extent it was over and done with early in the new century. Almost all of the pharmaceutical companies got into a feeding frenzy (this was on the back of the dot coms) and everyone seemed to think that this would be the solution to finding new drugs.

There were a very large number of businesses set up and of course the company that Craig Venter was associated with (Celera) was one of them. At the end of that process he had about a billion dollars in the bank from investors who were very excited about what this was all going to mean.

Now, what will happen in the near future as a result of the genomics revolution is that almost certainly we will get a read out as to the areas that we need to pay attention to (in health terms) as a result of our genetic profile. But even if a woman has a gene that predisposes her to breast cancer, it doesn’t mean that she’ll get breast cancer. It means that she’s got a gene that predisposes her to it. This is a huge difference and one that was not appreciated at the beginning of genomics.

So genomics is useful in a predictive sense, but it has been less useful in helping to find new drugs, and hence in helping to build commercial outcomes. Virtually all the genomics discoveries and investments happened in the US. There were a very large number of cashed up genomics companies that began to realise the product drought and so the issue became “Well, what do we do now?” What was needed was to work on the functional stuff (proteins) as these are where the new drugs or drug targets will come from.

Proteins do virtually all of the heavy lifting in the cells. They’re very complicated molecules. They’re traditionally very hard to work on and back in our laboratory in Macquarie University in the 90s, we had been working on proteins in a very slow and tedious way. We had about ten PhD theses written on characterising one molecule.

Through the period (1990s) when the genome was being deciphered, people were saying to guys like us, “You know you’re history guys, you’re doing it too slow, it’s all past, it’s all going to get sorted out without the need to work on proteins…the solution will be genomics, plus informatics, jumping over the hard stuff.”

But the reality is that proteins are complex and they also get all kinds of decorations. They get sugars attached to them, fats get added, phosphate groups are put on etc. None of this can be unravelled from the DNA alone, nor can you easily tell which cells contain what proteins.

So we knew that the answer was in proteins, and I think today that the consensus is that new drugs will come from studies on proteins. In fact many of the new drugs are proteins themselves. We are already seeing that diseases are being cured by giving people proteins or targeting defective proteins with small molecules.

What wasn’t working was how to do protein science fast. We spent a long time as university academics trying to work out how to do it fast. It is good to look back on that time and know that we are past it. My team decided that protein chemistry was really boring and if you tried to interest a big pharmaceutical business in protein chemistry, they didn’t want to hear, they just wanted to hear genomics. So we had a competition in the group to redefine our field of science.

One of my students (Marc Wilkins, now Executive VP, Discovery and Informatics at Proteome Systems) came up with the term Proteome. We thought it had a good sound to it. What we meant by the term ‘proteome’ was industrial protein science. That was way back in 1994. Now people are happy to focus on this new technology and the term ‘proteomics’ is in the dictionaries and is the subject of many biotech meetings (both scientific and commercially focused).

Then what happened was Craig Venter said something like “I’ve done genomics and now I’m going to do it again for proteomics.” Smart guy but he only lasted about six months. To his credit, he quickly understood that the barriers to entry in proteomics are much higher than they were in genomics. At that time, there was massive investment in proteomics, because there was a lot of cash around and people thought they could buy their way into a new technology.

If we look at the way transforming technologies get established, what tends to happen is you have a few hundred small businesses emerge and out of those few hundred small businesses you get a small group who survive. The idea of those cashed up was to short circuit this process by dominating with a mountain of cash…it did not work in proteomics and we are seeing the proof that difficult technologies require more than cash as the formerly cashed up proteomics companies struggle.

In the meantime, we were still at the University and we got major facility funding from the Federal Government. We learnt a lot about politics along the way. We told the government that proteomics is going to be a big deal even before anyone knew what proteomics was. Due to some very enlightened people in Canberra we eventually clawed our way into funding under the Major National Research Facilities Program. So we set up a facility at Macquarie University (The Australian Proteome Analysis Facility (APAF)).

This was the first major government facility in proteomics in the world. In the way of most government funding programs, we only got a partial solution. We got a building and we got really nice gear but we did not get staff salaries. So while I had a chair at the university, all of my staff, which I’d been carefully building up over a decade, were on soft money. It was a huge struggle to work out how to pay the bills and it became pretty obvious that this was not going to work. What was going to happen was the whole thing was going to fall apart as all of my guys would be picked off by attractive job offers, particularly from the US.

So at that stage our team got together and said “Let’s do it the way they do it on the West Coast and spin out”. That’s what we did. So we set up a business called Proteome Systems on the back of a deal with Dow AgroSciences, one of the world’s biggest agricultural biotech companies. We asked Dow to give us some difficult proteomics problems that they had not been able to solve. We said: “We’ll have a crack at them with the technology we’ve been developing”.

We were fortunate to be able to structure in that initial deal the means to build our whole business. So for the first 18 months of Proteome Systems’ existence we had no venture capital investment. We cash flow-funded the business, which was a great way to start. This was a very good discipline for running a business in a very lean and tight way. We understood that the technology was really quite inadequate to do this stuff but we had the expert people to make up the future. We still have probably the only group in the world that covers all of the territory of proteomics efficiently in an expert way.

We were very focused on coping with very large amounts of disparate information and bringing it together. That got us to appreciate how old fashioned the life science sector is in handling information. We paid a lot of attention to this. We had a very small team and they had a small set of Intel boxes to build an information management system that could handle all of the data. They also had to build the tools for manipulating the data, because a lot of the data comes out digital and you have to make sense of it. We achieved a lot and we had a nice package that we were proud of.

However, by this stage we were still only about 50 people. We wanted to approach big pharma with our solution but we realised that there was no way a small Aussie bio-tech company could get into a Pfizer or a Glaxo and convince them to put their lives on the line with our IT solutions.

So we went to IBM with a business model and said our goal is to sell an integrated technology solution in proteomics with IBM technology as the backbone. We got modern about how you can do life sciences. By partnering with IBM we could access all of the products that IBM uses in the enterprise computing space. Can you believe that people in life sciences labs have instruments holding huge amounts of valuable data, the hard disk crashes and you have lost everything? This is unbelievable but it happens in virtually every life science project in the world at the moment.

Our partnership with IBM is perhaps the best known example where we were able to move high up a relationship tree with a big corporation because of our ability to understand the difficult territory of proteomics. We have a global strategic alliance with IBM, their highest level partnership. They have about 100 throughout the whole of their business.

Companies like SAP and AT&T have this sort of alliance with IBM We co-market with IBM on our solution and it also gives us – we’re a tier one reseller - a lucrative business selling IBM equipment. I think we’re one of three companies in Australia that have such a status with IBM.

As we continued to build the business, we found that our goal was a bit different to quite a lot of our competitors. We did not assume that there would be a pile of cash available forever.

Therefore, we were very product focused because if you pay the bills, you pay the bills! Some of our competitors worked on highly technically complex solutions that were far away from products. While we have elements of our business that are very forward looking, we made sure that we got products to market fast.

We have very robust technology that we’re now able to sell. We’re partnered up with a number of big companies, so we haven’t set about having a big sales and marketing operation. IBM has 80,000 sales reps, so our problem is how to deal with an enormous sales force, rather than how to afford a sales force. This translates into other areas of our business where we have major partnerships with market leaders (eg Sigma Aldrich, Millipore, Shimadzu).

Basically what we’re looking to do, as happened in the DNA arena, is become the business that dominates the technology provider space. You do really well out of that if you get in early.

I think we are well positioned to be major technical and informatics providers in proteomics. I don’t think it’s a great long term business because it eventually becomes tough but there are a lot of good times along the way. It’s a great way to build a cash flow positive business and we’re well on the way towards being in that position.

Rapidly evolving is our diagnostics business where we’re looking at diagnosing disease and following its treatment or prognosis. After that, we are heading to be a drug company in our own right, and we have some very interesting developments in that area at the moment.

In conclusion, a few words on our location. Why are we still in Sydney when the conventional wisdom for Australia biotechs is often to move to the US as fast as possible? We like it here and we have never felt that being headquartered in Sydney is a disadvantage. For several reasons we feel that we are advantaged by being located in Sydney.

My reading of the history of Australian companies that try locating headquarters in the US is that it is pretty average. It is true that there are very few companies that have been successful. A company like Resmed springs to mind, but Peter Farrell built that business and he could have done it here had he chosen to do so.

Companies get built by true believers, not by bankers. Having said that, we know we have to be in the US and we have been there for several years, located in the hottest part of the US biotech world, in Boston. You have to have a presence, but you don’t need to be headquartered in the US.

The US made a massive investment in genomics and they’re still trying to recoup that investment. So there’s a huge amount of focus in the US still on trying to work out how to make money from DNA. Virtually all of the investments in the DNA space happened in the US, so there’s only country in the world that’s really suffering the big genomics investment.

Asia on the other hand is determined to make its way in the world, and it is not hindered by having invested in genomics. The good news is that the public sector effort in genomics eventually came through and this meant that the DNA based information became available free to corporations.

Proteomics companies use genomics information, but they don’t need to license this from genomics companies. This means that you can start afresh in Asia, with a proteomics investment, and that is what is happening. There is no baggage and you can just get on with it.

We are in great shape, dealing in our own time zone, with countries with thriving economies determined to be big players. Australians are well regarded in Asia and our partnering opportunities in countries such as Japan, The Peoples’ Republic of China, Malaysia, Taiwan, Korea, Singapore, India, are outstanding. Japan is a whole other story in itself. Proteome Systems has a joint venture with a major Japanese company (Itochu Corporation).

Together we have established Proteome Systems Japan to drive our business in Japan. We also have a big partnership with a major instrument group, Shimadzu. I could talk about other relationships throughout Asia, all of which leverage off our academic credentials and our ability to do business in the proteomic space.

The future is bright and we can do this from our time zone. The US and Europe are big markets, and we will have a significant presence in them, but Asia is the future and that is where we intend to do really well through partnership.

QUESTION TIME

Questions for Keith Williams

Can you tell us a little about your experiences with venture capitalists?

We got going without venture capital and so we avoided a group with money but lacking detailed knowledge trying to tell us how to run our business. In our space, even the expert technologists have trouble (eg Craig Venter did not make it) so how could you expect a VC to cope? The traditional VC dogma is that technologists can’t run businesses, and my dogma is that you don’t have a business if you don’t have technologists in significant decision making roles.

Whenever you start a business in this space, it never goes the way you think and you have to be flexible and knowledgeable. I am not convinced that the traditional venture capital model works in these emerging biotech areas, unless you are prepared to have massive attrition (which happens in the US).

Notwithstanding the above, we have had tremendous support from the Queensland Investment Corporation who have a longer term perspective on developing a biotech business.

I had an interesting conversation recently with a German biotechnologist who runs a biotech incubator in Munich about his experiences. The problem, he suggested, with VC was that the classic VC investor would come in with a small amount of cash, insufficient to make much progress.

The VC strategy involves going through several rounds of capital raising, so that the founders end up with only 1-5 per cent of the business. His view was that this is insufficiently motivating and leads to lack of interest in the founders because they don’t have enough incentive.

His incubator seeks to redress this by getting companies started so that when they raise capital, they do so without excessive founder dilution.

There are clearly scientists who don't cut it in terms of understanding the business and they need help but it makes no sense to marginalise them completely. The founders still own 80 per cent of Proteome Systems - and our motivation is very different.

What should biotech companies do right now?

The routine model is to define a product and then get out and sell it. Often this is a very skinny business, with no margin for failure. If you take our example, we are great at defining products and building them but we have little experience of selling.

Sales is not the skills base of most biotech companies at inception. I am not convinced that you have to immediately transform the emerging company into a sales and marketing operation. We leverage and outsource sales and marketing, leaving us to do what we do really well, giving away some of the good times, but getting a big reward nevertheless.

What does Australia need right now to make biotech long term and sustainable?

It would help to have a university system that was not being wrecked. I think we are in terrible trouble with our intellectual base at the moment and it is astonishing to me as I travel around the world that Australia seems to be the only country taking money out of our tertiary system and relying on individuals to cover the most basic costs of running the system.

This is a huge problem. Australia has benefited enormously through past investment in its university system and we are living off the fruits of that investment.

We don’t cope at all well with supporting emerging export orientated businesses as they move to commercialisation. There are some quite good schemes (eg START) supporting R & D, but after that it all falls away. If you actually make a product you become almost a leper….you can’t get loans from the bank as a start-up etc etc.

On the positive side, despite all the hype about areas to our North eg Singapore, the Australian biotech sector is by far the most robust in our region.

Venue