Sunday, October 4, 2009

Barcelona vista

Here are some thoughts about the second stop on our trip, Barcelona. We visited with Jaume Bertranpetit, and I gave a lecture to his group at the University of Pampeu Fabra. The group includes many other fine scientists in a beautiful new building, and we met a few who we had not known before. We had both previously been there as part of the PhD committee of one of Jaume's and his colleague Francesc Calafell's students. It was great to see these long-standing friends.

We were there during the weeklong festival of La Mercè, honoring one of the patron saints of the city, so it's only fitting to include a photo of some of the street theatre we saw. These guys were on flexible poles, swaying to music, an act we'd never seen the likes of before.

Jaume's group is probably the leading light of Spain in genetics, if not all of southern Europe, and has been very productive in front-line human genetics for some time. Our familiarity is with their work on human variation and history, especially of Europe (e.g., the reasons for the distribution of alleles associated with cystic fibrosis). They also do biomedical genetics, and are quite aware of the problems with GWAS and related ideas of how to map and deal with the genetic component of human disease. I talked on that subject, explaining how evolutionary perspectives show why we are finding what we are finding: that many important and common diseases that seem to have substantial heritability are not yielding to large-scale association mapping efforts. One member, Sergi Valverde, is a computational biologist trying to wrestle with the network approach to the problem.

We learned of a paper by Jaume and others in BMC Genomics, that looks at the degree of genetic difference in the 'isolate' populations compared to other 'main' populations. There it's reported that while the isolates do have some genetic differences from the surrounding populations, they are not really as different as their language.

A classic example, though not one reported in this paper but in another unpublished paper by Dr Bertranpetit is the Basques. Their isolation was long argued based on their language (apparently ancient and unrelated to what is spoken by surrounding populations in Spain and France).

This is relevant beyond the long-standing interest in the history of the Basques, about which many classical papers in human genetics from the early to mid 20th century were written, or other individual isolates. After a few traits were found to be at high frequency in religious isolates (like the US Amish and Hutterites, in Finland and in French Canadians), the idea became widespread that the bottleneck in population size would simplify the genetic causal basis of otherwise more complex diseases, making them more mappable in the isolate.

This general isolate or bottleneck effect goes beyond the classical reason, that founder effects (recent founding by a small number of people, plus increased relative degree of inbreeding in the small descendant population) would by chance raise the frequency of some recessive traits to high enough frequency that they could be mapped and understood.

However, in the Basques as elsewhere (as Bertranpetit and colleagues note), the degree of bottleneck and the amount of subsequent gene flow with surrounding populations, the differences are far less than was expected. They are not proving to be the mapping bonanza that was hoped. Given the amount of variation that exists, the amount of population constriction needed to do what was hoped is far more severe than has been the case. Of course, isolates may have some disorders at higher frequency, or some very rare causal genotypes basically not found frequently enough elsewhere to be useful. But overall, isolates were over-sold.

In fact, the relative success in mapping in the Iceland population, which was touted in advance as having great potential largely to their isolate status, has proven to be due mainly to the availability of large genealogies, not isolate status. Even in Finland, where much of the hope for the usefulness of isolates was fueled about 15 years ago, and where the local population history can be reconstructed, there has not been a high yield of new genotypes found to be associated with common disease.

The Basque story shows that one has to be circumspect about ideas that seem plausible but can be accepted too uncritically. The history of the Basques is more interesting as history than as genetics. That's a subject too great to go into but a fine book, The Basque History of the World by Mark Kurlansky provides some good material, as does web searching, etc.

Besides seeing our daughter newly installed as a post-graduate violin student in Barcelona, we had a very fine time visiting with Jaume and others. It is a city worth visiting, and Jaume's is a scientifically vibrant group. In addition to his science, he is an effective politician when it comes to program building and resource acquisition. For the last few years he has been the very successful director of a program to recruit scientists to Catalunya, ICREA, which may still be looking for new people, in case any reader might be interested.


1 comment:

Francesc said...

Thanks for the post, you maed me blush! Just in case someone is interested, the Basque LD paper can be found at Eur J Hum Genet (Garagnani et al., still in press)