Q&A: John Hooper, CEO, Genizon Biosciences

Hooper: There are two groups of reasons. One reason is populations which had arrived from a small number of families tend to reflect the mutations that were brought in by the founders. If there are only a small number of (families) then you only have a small number of mutations. For instance, the BRCA1 gene which is involved in some forms of breast cancer has about 1,200 mutations world wide. There are only around a dozen in Quebec and only three of those are really important. So you have great­ly reduced numbers of muta­tions and if you are looking for any particular mutation that is causing disease it is a lot easier to pick it out from say 10 than it is 1,000. That's putting it crudely; it is much more statistical than that.

 

The second reason is called "LD" (linkage dis­equilibrium) which is the extent to which any two individuals in a population share chunks of DNA, much like the blocks that were studied in the HapMap Project. We've done a kind of HapMap Project on the Quebec Founders Population and we find that their blocks are significantly larger than the blocks in the general popula­tion and as a result of that we don't need to use so many SNPs to determine disease. The consensus is that in the general population you will need somewhere between 300,000 and a million SNPs to study in each person to be able to identify a disease gene and you'll need a lot more patients because it is not a homogeneous population. In Quebec we generated our Gene Map for psoriasis with only 60,000 SNPs in about 500 people.

Hooper: There is a very practi­cal reason and that is there are 6 million people in the QFP. So recruiting a thousand people and a thousand controls is not an oner­ous task. But if you are doing it in Iceland, and you want a thousand people, it is a thousand people from 280,000 where only about 2,800 will have psoriasis and some of them may not have the right kind of psoriasis. That is one practical consideration, I think another one is the age of our population which is 12 to 14 generations old. And that is somewhere in the middle of the sweet spot, so to speak. For instance the Icelandic population is about 20 generations old and that is quite a lot of time for new mutations to arrive and it is also quite a lot of time for intermarriage and while that particular popula­tion has been fairly well isolated, there is quite a bit of opportunity for that to happen in older popula­tions.

Hooper: I think so. But it is not just the QFP that makes us unique. There is a significant computing challenge involved here as well. People tend to gloss over it as merely something that computers can do, but there is a lot of soft­ware and a lot of considerations of the human genome and how it behaves, because you can't get into creating (these tools) until you gen­erate the data. It has taken us two years simply to figure out how to use our data. In a general popu­lation you are talking an order of magnitude bigger and there is going to be quite a potential com­puting problem there that I think people have underestimated.

Hooper: Well this is really diffi­cult. Because if we tell a potential partner what the genes are, then they have the information they need to move forward. So we are very keen not to contaminate any­body we are trying to sell a gene map to, in case they don't buy it and they are keen to avoid being con­taminated, as well. What we try to do is show what we have, together with what they have, in a theoreti­cal black box, through either an external consultant whom they trust and who will then make a report without revealing the iden­tity of genes. Or we provide the information to someone within their organization who is governed by the laws of their profession—a patent attorney is a good person—where one might lose a license to practice should any information be revealed.

Hooper: We want someone who can take our discoveries down­stream in a fairly rapid fashion. The last thing we want to do is license a discovery to somebody and then have it sit on the shelf. So we are looking for companies that are dedicated to the particular disease category we are looking at and is willing to take what we have and put a team on it.