Why sex?
Sex is something that most people think about on a minutely basis, but most don’t *really* think about it. From a macroscopic evolutionary view, sex doesn’t really make sense. The amount of effort that males put in to attract females, even outside the human species, is extreme. Sex is expensive in terms of time, effort, and stress. It would be much easier and less expensive from a reproduction standpoint to undergo binary fission, or drop pieces of oneself every time one wanted to reproduce.
Of course we don’t do that, and there are a few explanations as to why this is. (Because it’s “fun” doesn’t apply: that’s an evolutionary byproduct of needing to reproduce.) One hypothesis put forth almost 20 years ago suggests that sex evolved as a way to purge harmful mutations from the population. By shuffling genes “randomly” (mixing chromosomes is anything but random — any sociologist will tell you that), the harmful mutations would be concentrated into a few select individuals who would be weaker and less likely to reproduce, and therefore these mutations would be weeded out through natural selection. (more…)
Tackling malaria using distributed computing
Malaria caused by Plasmodium falciparum is one of the leading causes of death in the third world, especially among children. Becoming immune to all of the variations of Plasmodium malaria can take upwards of five years. The reason for this is because of the cloaking capabilities the parasite has evolved over the years. A process called epigenetic silencing allows the Plasmodium to express only one antigenic protein at a time. As there are about 60 genes that can be turned on and off, this means the body must learn to recognize 60 different forms of the same organism.
A distributed computing project out of France aims to tackle drug-discovery for Plasmodium-mediated malaria. Using software developed by the Fraunhofer Institute — the same people that developed the MP3 codec — the project narrows down the list of possible drug candidates to a select few which will be further analyzed by supercomputer.
The project, called Wide In Silico Docking on Malaria (WISDOM) to model 3D structures of proteins from Plasmodium to ligands: the chemical compounds that bind to protein receptors. The massive parallelism was achieved by assigning one ligand to one protein to each node on the grid. Computing the probability of a match can take a few seconds to a few minutes.
The project joins myriad other distributed computing projects in the life sciences, many of which are specifically drug-discovery efforts for diseases ranging from cancer to AIDS to anthrax and even Ebola.
Technorati Tags: WISDOM, distributed computing, grid computing, malaria, plasmodium falciparum
Revisiting the Easter Island mathematical modeling
Dr. William Basener, whose approach to modeling a societal collapse I wrote about a few days ago, left me some comments, but they got truncated for some reason, so I asked him to continue. He responded, and what he has to say is so extensive (not to mention interesting) that I opted to turn it into a feature article. I hope he doesn’t mind; it’s quite fascinating and readable.
I’ve always found the the behind the scenes stuff more interesting than the press release itself. Come on inside and check it out!
Q&A with Dr. William Basener on his Easter Island collapse model
Modeling the collapse of the Easter Island culture
William Basener, an assistant professor of mathematics has created the first mathematical formula to “accurately model [Easter] island’s monumental societal collapse.” I am skeptical of this claim because no historical records were kept by the islanders regarding what actually happened, so claiming something is accurate when it cannot actually be verified seems a little disingenuous to me. I don’t like things being presented as fact when they are not falsifiable. Nonetheless, this is a press release, so I suppose that such language is to be expected.
The story is interesting, though, because Easter Island might be an accurate microcosm of what could happen here on Earth in the next couple of thousand years, assuming we don’t all wipe each other out first.
Between 1200 and 1500 A.D., the small, remote island, 2,000 miles off the coast of Chile, was inhabited by over 10,000 people and had a relatively sophisticated and technologically advanced society. During this time, inhabitants used large boats for fishing and navigation, constructed numerous buildings and built many of the large statues, known as Tiki Gods, for which the island is now best known. However, by the late 18th century, when European explorers first discovered the island, the population had dropped to 2,000 and islanders were living in near primitive conditions, with almost all elements of the previous society completely wiped out.
“The reasons behind the Easter Island population crash are complex but do stem from the fact that the inhabitants eventually ran out of finite resources, including food and building materials, causing a massive famine and the collapse of their society,” Basener says. “Unfortunately, none of the current mathematical models used to study population development predict this sort of growth and quick decay in human communities.”
The reason that such models haven’t been developed is simply because population groups that live long enough in isolation to collapse are few and far between. This same collapse will happen to Earth eventually, since this is actually a planet with limited resources. This is one of the many reasons that space exploration is so important. Not for us here and now, perhaps, but certainly for our progeny, because a time will come when we do run out of resources here on Earth.
Dr. Basener plans to turn his formula to the Mayan and Viking populations next, and eventually hopes to modify his work to predict population changes in the modern world. He hopes that it will lead to the development of better population management skills so that future famines and population collapses may be averted.
Unfortunately for Dr. Basener, population “collapses” as a result of famine are relatively uncommon — they usually happen as a result of war: historically it has been common for one group to conquer another, kill all the men and children, and dilute their gene pool by raping their women. It is difficult to find significant population groups that remain isolated for long periods of time where they could implode under their own pressure. Humanity’s biggest threat in the near future isn’t imploding due to famine and mismanagement; it’s imploding due to something akin to nuclear winter. (Though now that the Cold War is over, this seems unlikely, though the possibility still exists.) In any event, I believe that Dr. Basener’s work would be much more interesting if it were applied to the world as a whole rather than just small populations within the world.
Update: There is a Q&A with Dr. Basener here.
The intricate hurricane cycle
By now, you’ve heard about what Hurricane Katrina has done to the Gulf Coast, specifically, to New Orleans. Many people have been wondering why there’s been so much hurricane activity in the last few years, myself included. There have been more hurricanes and tropical storms, and on average, they’ve been fiercer and more destructive. So what’s going on? NASA’s Earth Observatory has got an article up about the hurricane cycle.
Using mathematical models, Doctors Xie and Pietrafesa developed a way to analyze patterns relating to tropical storms dating from 1887 to 1999. By analyzing patterns, they predict that the current hurricane resurgence will continue for the next 10 to 40 years. Not a very exact (or encouraging) prediction, but it does indicate that if New Orleans is rebuilt the same way the old city was, it could very well be wiped out again by another powerful hurricane like Katrina. In the the prophetic words of the Times-Picayune, “It’s only a matter of time.”
Their model is based on four cycles, shorter cycles within larger ones. A long-term cycle contributes one or two tropical storms per year to the total number, and a short-term cycle contributes up to five per year for the North Atlantic. Since 1995, there has been a dramatic increase in hurricane activity, with storms like Fran and Floyd causing widespread destruction.
An earlier, July 20 study published in Science suggests that the recent increase in hurricanes and tropical storms is the beginning of a 20 to 50 year trend in hurricane activity. But Xie and Pietrafesa believe that their analysis technique, called Empirical Mode Decomposition (EMD) more accurate describes the patterns of tropical storm occurences.
The most energetic cycle - the one that shows the greatest variation in the number of landfalling storms between its peak and low-point - is one that lasts three to five years. That cycle, Xie says, essentially adds or subtracts one or two landfalling tropical storm events every year on the East Coast. The eight-to-12-year cycle can add or subtract an average of one and a half hurricanes per year; the 20-40-year cycle can add or subtract an average of about one-half hurricane per year; and the longest, 40-60 year cycle (similar to the cycle described by Goldenberg and his colleagues) can add or subtract about one hurricane per year.
The causes of these trends are not immediately clear, but global warming has been implicated, along with the oscillations of ocean currents in the Atlantic and Pacific Oceans. Among these oscillations are El Nino (the warming) and La Nina (the cooling) of these two oceans.
During the 112-year period for which the NC State researchers have data, an average of 3.23 tropical cyclones pounded the East Coast each year. During El Nino years, that number dropped to an average of 2.47 storms. North Carolina saw an overall average of 0.81 tropical cyclones annually, and 0.56 during El Nino years.
Jonathan has great coverage of more of the science behind how and why Katrina got as powerful as she did in this week’s Science.Ars. I will investigate the global warming angle on the hurricane season in the next few days. The scientific understanding of the ferocity of a given hurricane season is not fully understood, but headway is being made. Observing patterns is the first step towards explaining them, and right now it seems we’re on the upswing of a few of the observed hurricane cycles — possibly all four.
Chain emails and gas prices
Personally, I’ve not gotten the chain email that’s going around, but it seems to be quite widespread. I was reading the Freakonomics blog (an excellent read, by the way), and I came across the rebuttal of the chain email stating that if we all didn’t buy gas for a day, we’d ruin the oil cartel.
I remember reading a similar email years ago, and for some reason, I always wondered idly if it were true. Not having thought about it much, I sort of forgot about it. But with all the insanity in the US surrounding the rising gas prices, it has resurfaced, more popular than ever. Officially, today was to be the day that no one bought gas. I’m sure that this didn’t happen for myriad reasons, the main one being that despite the size and saturation of Internet access in American households, the Internet is still fairly new, and newspapers and television still reach a wider audience. Even massive websites, like Slashdot, only reach an average of ~2000 people per million, or 0.2% of the population. So even if a place like Slashdot had posted it, it still wouldn’t have made that big of an impact. Of course part of the slashdot effect is that many other, smaller websites pick up whatever is posted, further disseminating it to the masses. Supposing that doubled the number of viewers, that would still only be 0.4% of the population.
I think you can see where I’m going with this. Chain emails are futile for getting anything changed, and so is posting on websites that have niche audiences… even those sites that would be considered huge in absolute number of visitors. It is virtually impossible to reach the majority of the American public because they all have different interests and tune into different things.
Alas, I’m rambling. Back to gas prices and bringing down the oil cartel. If you read the post on Freakonomics, Levitt systematically disassembles the email piece by piece, calling into question the figures bandied about, among other things. I largely agree with him, except with one minor nitpick.
If nobody buys gas today, but everybody drives the same amount, then it just means that we either had to buy more gas in anticipation of not buying any on September 1, or that we will buy more a few days later. So even if you believed this would take a $4.6 billion dollar bite out of the oil companies that day, consumers would hand it right back over. If this was “No Starbucks coffee day” it might have some chance of mattering, because people buy and drink Starbucks coffee the same day, so a foregone cup of coffee today may never be consumed. But this is not true of gasoline, especially if no one is being asked to reduce gas consumption. All you will get is longer lines at the pump the day after.
Even supposing the impossible were to happen — everyone opting to not buy gas on September 1 — I don’t necessarily think that the days before and following the chosen date would necessarily completely make up for people not buying gas on that day because people will associated not buying gas with not driving, despite assertions to the contrary. That’s my only minor nitpick, and I would go so far as to extend Levitt’s paragraph quoted above a little bit further. For simplicity’s sake, assuming that roughly the same amount of gas is sold every single day, seven days a week, only 1/7th of the gas-buying population would have filled up today anyway. So at most that would have been a 14% impact to the bottom line for the week. In the big scheme of things, this little blip on the radar isn’t a whole lot when the year as taken as an aggregate, further illuminating how silly the idea of bringing down the oil cartel by not buying gas for a day truly is. Of course, everyone knows that the real solution to the gas problem is greater efficiency, driving less, and finding alternative fuel sources.
Why are gas prices rising, then? That’s the question I’d like to see answered. I have a few thoughts on why, but I know I don’t have the whole story. Nonetheless, I’ll share my thoughts with you.
With Hurricane Katrina tearing NoLa to shreds (donate!), I’ve been watching the news more than I normally do, and the hurricane coverage is typically followed by the next biggest story: soaring gas prices. Indeed, prices jumped about a quarter literally overnight here in New England. Interviewed on the news are lots of understandably angry residents. Some of the comments are… interesting to say the least. One man complained that the president and the government should do something. Another person complained that the government should lower the prices, as if they somehow magically controlled the pricing. I felt a great deal of dismay after watching these interviews, because it was depressing to see how ignorant a large section of the population really is. (As a side note, they also interviewed a Canadian, and he had the most balanced view of anyone they interviewed, noting that we Americans actually have it pretty good. Does that say something about most Americans, or was he just an astute Canadian? That question I leave for you to ponder.)
With respect to the government, about the only thing they can do is decrease its oil reserve to temporarily increase the domestic oil supply. Naturally, the laws of supply and demand determine the prices of nearly everything, and increasing the available stock will decrease the price at the pump. Now, just about everyone knows that OPEC controls the majority of the world’s oil supply, and that they have been known to artificially decrease the supply (by pumping less oil) to increase the price of a barrel of crude. This time, though, they probably aren’t doing this. Instead, it seems as though China might be to blame. As China modernizes, their demand for oil increases, and it seems as though OPEC might be having a hard time keeping up with the demands of both the East and the West. As a result, prices are rising because this time, the scarcity is real, leading to an increase in price at the pump.
I know that this is only one piece to a very complicated puzzle, but I think that it is something that citizens should bear in mind the next time they start to complain about the price of gas (or anything else): it’s about supply and demand — artificial or real is irrelevent because the net result is the same. Without getting into very complex shades of political gray — Halliburton, the war in Iraq, taxes on oil — the government does not control everything, and as such, they should not be held responsible for every little problem that causes Americans as a collective, to gripe. There’s nearly always another side to story, and the mainstream media should do their best to educate rather than sensationalize. (Yes I crack myself up sometimes.)
Autonomous language learning
I was hoping David would write this up, but I’ll do my best. He said he was going to attempt to get the fulltext of the paper. I hope he does and decides to write an article on it. (*nudge* *nudge*)
Researchers have created a program which can infer the rules of grammar by being fed text in just about any language. It can do the same for music, gene sequences, and proteomics.
“This is the first time an unsupervised algorithm is shown capable of learning complex syntax, generating grammatical new sentences and proving useful in other fields that call for structure discovery from raw data, such as bioinformatics,” he said.
Unlike previous attempts at developing computer algorithms for language learning, the new method, called Automatic Distillation of Structure (ADIOS), successfully identifies complex patterns in raw texts. The algorithm discovers the patterns by repeatedly aligning sentences and looking for overlapping parts.
For example, the sentences I would like to book a first-class flight to Chicago ,I want to book a first-class flight to Boston , and Book a first-class flight for me, please may give rise to the pattern book a first-class flight — if this candidate pattern passes the novel statistical significance test that is the core of the algorithm.
If the system also encounters the sentences I need to book a direct flight from New York to Tel Aviv andI would like to book an economy flight , it may infer that the phrases first-class , direct and economy are equivalent in the context of the new pattern. “Because such equivalence sets can contain other patterns — in turn containing further patterns, and so on — the resulting body of knowledge grows recursively, as a sort of forest of branching trees of possibilities,” said Edelman.
He added, “ADIOS relies on a statistical method for pattern extraction and on structured generalization — two processes that have been implicated in language acquisition. Our experiments show that it can acquire intricate structures from raw data, including transcripts of parents’ speech directed at 2- or 3-year-olds. This may eventually help researchers understand how children, who learn language in a similar item-by-item fashion and with very little supervision, eventually master the full complexities of their native tongue.”
The word “unsupervised” sounds dangerous, almost Skynet-like, but I think we’re still quite a ways from that. I bet such an algorithm could be adapted to scan for malicious code, viruses, or almost anything else. I could see implications when it comes to anti-terrorism and plain old espionage. How about a computer that could realistically “learn” various languages and be taught to scan for certain types of conversations? A higher-level, more abstract understanding of language could certainly flag fewer false positives and be more accurate than keyword density scans of random conversation. Of course, I’m just speculating on how agencies like the CIA, NSA, and FBI actually do this information processing; they probably have other, better ways of doing it.
Getting more pragmatic, I bet this could even be used to build more capable spam filters and firewalls, since the algorithm seems blind to written language and binary, it could conceivably scan and learn anything, and check for any sort of anomaly.
I would think that the next step is moving from passive information processing to active interaction with another human being or problem: how long before a program using this algorithm can solve mathematical word problems, or perhaps even pass the Turing test? I think this is a huge step in both of those directions, but I remain leery of the theoretical capabilities of such an algorithm being used for unethical purposes.