Death of the World's Oldest Woman

According to the BBC, Maria Esther de Capovilla died on Sunday at age 116. As usual, her relatives ascribe her longevity to any of a variety of healthy practices - eating well, not drinking or smoking, religious devotion - but it really all comes down to chance. Someone has to be the oldest person in the world at any given time, after all! Not that I'm decrying the value of a healthy lifestyle; one of these days I'm going to give it a go myself. But humans are storytellers; we just can't let facts exist without enmeshing them in a web of conjecture. This is how it should be.

I was particularly struck by the information, mentioned in passing in the BBC article, that she spent the last 20 years living with her daughter and son-in-law. I'm sure she was a lovely person, but you can't help feeling for a guy who marries the woman of his dreams and ends up living with the world's oldest and most enduring mother-in-law!

The oldest person ever documented is, of course, Jeanne Louise Calment, who died in 1997, aged 122. It looks like 120-odd is about the human maximum, at least without genetic manipulation (and we have no idea how to do that yet). Assuming that most women marry in their 20s, their mothers are likely to be in their 40s, with 80 years at most ahead of them. So sons-in-law worldwide have an upper limit to their years spent housing Mother!

Aging and Cancer

In a previous life (well, this time last year) I was working on the genetic networks underlying cancer. In this incarnation I'm working on the genetic networks underlying aging (at least, I will be when we've finished building the databases, identifying data sources, obtaining and integrating the data, constructing the networks and having a couple of celebratory pints at the Trent House...) And it's becoming more and more apparent that there really isn't a lot of difference between the two networks.

From one persepctive, of course, that's a trivial observation. All human cells contain the same set of genes, so the Ultimate Network is the same. But of course different genes are expressed in different types of cell, so different subnetworks are actually active in different cells and at different times. And it turns out, perhaps not so surprisingly, that the sets of 'cancer' genes and 'aging' genes[1] have significant overlap.

One of the most downloaded articles from Nature was an editorial called 'Ageing: The price of tumor suppression?'[2], reporting on research published by Tyner et al. in the same issue [3]. The Tyner team found that mice with mutations making the gene p53 (the 'guardian of the genome') more active appeared to have accelerated ageing. p53 is a tumor suppresor gene, and you'd expect that making it more active would be beneficial, in that it would protect against cancer, but the hapless beasties ended up with "reduced longevity, osteoporosis, generalized organ atrophy and a diminished stress tolerance". They reason (the researchers, not the mice) that ageing may be a side effect of the natural safeguards that protect us from cancer. Which would mean that combatting ageing by genetic means might just lead to a higher cancer risk. It's like the question all kids seem to ponder: would you rather die by drowning or by shooting?

Confusingly, there's a recent article in Science by Pinkston et al.[4], who found that in C. elegans, everybody's favourite worm, mutations that increase the lifespan also inhibit tumor growth - exactly the opposite of what you'd expect from the mouse models! I'm not sure how applicable this finding is to humans, though; worms don't usually get cancer, so they were genetically engineered to produce tumours, which means that the tumours are probably fairly unlike human tumours. Still it's an interesting finding, and highlights how little we know about the genetics of cell growth and death. Looks like I might be in a job for a while yet!

[1] Of course, there's no such thing as a gene for cancer, or just about anything else, media reports to the contrary. I'm just being lazy and using the usual shorthand for "a gene whose function (or malfunction) impacts the cancer phenotype.
[2] Ferbeyre, G. & Lowe, S. W. (2002). Ageing: The price of tunour suppression? Nature 415:26 - 27.
[3] Tyner, S. D., Venkatachalam, S., Choi, J., Jones, S., Ghebraniousk, N., Igelmann, H., Lu, X., Soron, G., Cooper, B., Brayton, C., Park, S. H., Thompson, T., Karsenty, G., Bradley, A. & Donehower, L. A. (2002). p53 mutant mice that display early ageing-associated phenotypes. Nature 415: 45 - 53.
[4] Pinkston, J. M., Garigan, D., Hansen, M. & Kenyon, C. (2006). Mutations That Increase the Life Span of C. elegans Inhibit Tumor Growth. Science 313:971 - 975.

The Big Adventure: A Retrospective

Well, I'm back in Newcastle. It's about 17 degrees, grey, and has rained almost constantly since we got back; something tells me summer is over! To add to my woes someone cloned my debit card in Fortaleza and cleaned out my bank account; I forgot to pay BT before I went overseas and my phone (and hence internet connection!) was cut off; and my motorbike was stolen from the parking lot opposite my flat and vandalized. So the homecoming hasn't been quite as cheerful as it could have been. On the plus side I have a bathrobe with a hotel crest, a bottle of Brazilian rum worth 6 Real (about 2 pounds) and a headful of inspirations, scientific and otherwise. It's been quite a month.

The contrast between the two conferences I went to is fascinating. CEC is an IEEE conference, and as such is full of engineers and computer scientists. In the biology tracks the biological questions are of interest, naturally, but the focus is on algorithms: on turning raw biological data into interesting information. ISMB is dominated by lab-oriented biologists, even though it's the official conference of the International Society for Computational Biology. I got the distinct impression that many of the attendees see bioinformaticians as the handmaidens of "real" biology, rather than as biologists in our own right. There was even a keynote presentation on "Why bioinformaticians need experimental biologists". This is an attitude with which I'm all too familiar, but it's disappointing to see it amongst this community. It needs to be stated that there is some biology-real, honest-to-God biology-that can only be done using a computer. Clearly, we need the lab dwellers and their data, and equally clearly there is a need for computer support for lab biologists, but computational biology is a valid field in its own right. Of course, it's early days yet, and computational biology has yet to prove itself. We need some major breakthroughs, and I think it will be a couple of years before that happens.

OK, off the soapbox and back to doing some of that biology! Or possibly to devising appropriate responses to punk who trash other peoples' motorbikes... Matt's suggestion of red hot knitting needles has a lot going for it.

ISMB 2006 – Day 2

An interesting day, particularly because it had a couple of more theoretical, evolutionary biology type papers, which I really enjoy. It’s good sometimes to take a step back from the nitty gritty detail with which we work every day and think about the bigger issues in biology. I sometimes suspect I’m an evolutionary biologist manqué, but I really believe, deep down, that all biologists should be.

OK, so – first notable talk (of those I saw!): Christina Chen on ultraconserved elements in genomes. I’m a bit wary of this whole “ultraconserved” business; it seems to me that we really don’t know enough about comparative genomics to state definitively that long stretches (200 appears to be the cutoff) are really completely unchanged between species. And Christina’s research seems to support that. They’ve found that putative ultraconserved regions actually are polymorphic between individuals (although at significantly lower rates than expected) and that there’s no evidence for strong selection on these alleles, because they’re in Hardy-Weinberg equilibrium[1]. And the conclusions? I think we have to be very careful in our interpretation of the meaning of “ultraconserved” regions in genomes. OK, it may not be a brilliant insight, but there’s been a lot of hype in this area lately!

The final plenary was also excellent: Mathieu Blanchette (the Overton Award winner) on “What mammalian genomes tell us about our ancestors and vice versa”. He covered three major issues: predicting transcriptional regulation, the characteristics of CRUNCs (coding regions under non-coding selection), and his Ancestral mammalian genome reconstruction project, in which they’re attempting, with what looks like considerable success, to reconstruct the genome of one of the common mammalian ancestors from the mammalian radiation of 75 million years ago [2]. It’s cool not only in it’s own right, but because it allows comparison between modern mammalian genomes and their common ancestor: as close as we can get to evolution in action.

All in all, a good day, rounded off by a great dinner at (by Fortaleza standards) an outrageous price. Well worth it!

[1] Not that selection is the only factor affecting H-W equilibrium; when I was an undergrad we used to drink toasts to Hardy and Weinberg because the conditions required for their equilibrium are: 1) No selection; 2) No migration; and 3) Random mating. Guess which one we used to drink to?

[2] The Boreoeutherian ancestor, to be as precise as my notes allow. And no, I have no idea what it was.

ISMB 2006 - Day 1

We dragged ourselves away from the beaches and the bars today to catch a bus to the Fortaleza conference centre at the crack of dawn. There are about 800 attendees, I gather, and they all arrived at around 8am on half a dozen buses. Eight hundred computer scientists who had been deprived of internet access for periods of up to 60 hours (I'd like to hear that story sometime!) arrived more-or-less simultaneously at a venue with wireless access. Which promptly collapsed in a wheezing heap. Matters seem to have improved this afternoon, however; maybe half the delagates are on the beach?

I've been in the Systems Biology stream all day, and it's been excellent. Lots of network talks, most of them addressing the sort of issues that our group wrestles with every day: not enough data, too much data and ill-formed data (usually all at once); the need to make what we know are unrealistic simplifications; computational feasibility. The picture I'm getting is that the field really has advanced over the last couple of years, but there's still an awfully long way to go. The emphasis on biological data (although not necessarily biologically important questions!) is good, too. The origins of the society's journal, Bioinformatics, as Computer Applications in the Biosciences is gradually being lost, although it's taken time.

The second plenary this afternoon looks promising. But maybe I'd better pay attention to the talk I'm currently in...