This has to be one of the funniest videos of the year. I love nerd humor...
The lyrics, for those who might be a bit confused:
There was a time when to amplify DNA,
You had to grow tons and tons of tiny cells.
Then along came a guy named Doctor Kary Mullis,
Said you can amplify in vitro just as well.
Just mix your template with a buffer and some primers,
Nucleotides and polymerases, too.
Denaturing, annealing and extending.
Well, it's amazing what heating and cooling and heating will do.
(chorus)
PCR, when you need to detect mutation.
PCR, when you need to recombine.
PCR, when you need to find out who the daddy is.
PCR, when you need to solve a crime.
It's been quite awhile since I sat down and read a biology paper. But today I found a review article that piqued my interest. It was published a few weeks ago in Science titled A General Model of Prion Strains and Their Pathogenicity (pdf).
I've always been interested in prions, which are self replicating proteins that causes diseases commonly known as "Mad Cow". We know very little about these proteins and they apparently form a totally new class of infectious agents (more common pathogens are bacterias, viruses, or protozoas).
What's interesting is that prions are made only of proteins (no DNA or RNA) yet they self-replicate. The current model is that PrPC (cellular prion protein) is recruited and aggregated by PrPSc (a related protein made from PrPC and other monomer units). Then as the polymer grows, at some point PrPSc cuts the prion polymer to yield two PrPSc. This is how the protein self-replicates and achieves exponential growth.
There are many strains of prions. Some strains, such as the one that causes mad cow disease, readily transmits to other mammal species (such as rats or humans). Yet some other strains do not jump across species. The proposed model for this is that some PrPSc are compatible across different strains of prions, which allows compatible prions from another specie to replicate. But when there are no overlap in the compatible PrPSc among the different prion strains, then this becomes a barrier to transmission.
Lastly, there is evidence to suggest that PrPSc themselves are not toxic and are not the proximal cause of disease symptoms. Instead, during the reaction that converts PrPC to PrPSc, an intermediate by-product protein, PrPL is sometimes created. It is theorized that the accumulation of PrPL causes the onset of symptoms and ultimately result in death.
It took me awhile to plow through this article (it was pretty dense, at least for me) and I fell asleep halfway through it. But this doesn't mean it wasn't a readable, informative, and interesting article. If you're adventurous, you should give it a read and let me know how far you get.
I'm way too busy to write about anything sensible, so I'm just going to post this video. It's a BBC documentary on this lesser known phenomenon called global dimming that I happened to stumble upon on Google Video.
It's a bit long, almost an hour. So you probably want to go and look at the big version. But I think it's well worth it. The overwhelming feeling coming out of it is that we're basically screwed (and not in a good way). However, these BBC science video do try to make things more dramatic than they are in real life.
But the science seems to be solid. So I tend to want to give it the benefit of the doubt.
Yesterday, as I sat in the front row listening to Vinod Khosla talk about his adventures in renewable energy, I felt my throat become scratchy. This usually means I'm coming down with a cold or something. Graduate school has absolutely shot my immune system to hell.
Today I went to Walgreen's, because the Long's Drugs across the street from my apartment decided to close forever yesterday (how convenient). I recently heard from some people that they take high doses of vitamin C to prevent colds. One particular product comes to mind: Airborne. Apparently it is the fastest selling dietary supplements in retail history. Naturally, I bought some, in the hopes that it will boost my immune system so that I actually don't get sick (or it passes quickly).
But, scientific evidence is still not clear (read: I'm not convinced). This PLoS Medicine review summarises previous clinical trial results. This sentence about sums it up:
The clinical significance of the minor reduction in duration of common cold episodes experienced during prophylaxis is questionable, although the consistency of these findings points to a genuine biological effect.
Basically, vitamin C appears to help cold sufferers, but only marginally. It might work better at restoring your immune system to full capacity after stress or harsh environments, but it doesn't seem to do much as a prophylatic. Science is kind of weird when it depends on statistical data to "prove" anything. I forsee the two sides arguing forever on this issue, unless we can actually pinpoint the molecular mechanisms of how vitamin C interacts with the immune system. Anything short of that, we can always fudge the numbers to show one side is better than another, regardless of which side.
In the mean time, I'm taking vitamin C, just to try it out.
Yesterday, I learned something about myself. Anyone who knows me, know that I don't drink alcohol. And if you really know me, you've seen me drink. The explanation I give to most people is that I go from stone cold sober to throwing up in half a beer (give or take). So finally, yesterday, I dug around and found some information on the internet about my condition.
My condition is known as alcohol flush reaction or more commonly known as Asian Blush (I personally dislike the latter term). It affects about half of the East Asian population. To those afflicted, drinking even moderate amounts of alcohol will cause the face and body to burn red and heart rate to soar. And for me in particular, I get nauseous which naturally leads to vomiting.
As far as molecular biology is concerned, this condition is caused by two gene mutations. First, ethanol is decomposed into aldehyde by alcohol dehydrogenase (ADH) enzyme. Many Asians have a mutated form of ADH that causes this reaction to occur 400% faster than normal. This sounds great right? Well, no. It turns out that aldehyde is more toxic than ethanol in the blood. But we're not done yet. The second reaction, aldehyde gets converted into acetic acid (vinegar) by the enzyme aldehyde dehydrogenase (ALDH). About half of Asians have a mutated form of ALDH that functions an order of magnitude worse than the normal enzyme. So these two put together ensures that when I drink alcohol, my body fills up with aldehyde. It is aldehyde that causes those adverse reactions causing me to throw up violently. I suspect that I have the homozygous mutated form of the gene, thus leaving me completely unable to handle alcohol (those with heterozygous mutations have it much easier). It's not that I get drunk on ethanol really fast, it's just that I get poisoned by aldehyde.
Unfortunately, there is currently no cure. There are a few ways around this, but none of it is recommended by current medical practice. For example, taking heart burn medication such as Pepcid AC somehow causes the first reaction to take place slower. Therefore less aldehyde is produced initially. However, this is somewhat unsatisfactory since aldehyde is still metabolized at the slower rate in the second reaction. Since aldehyde is much worse than ethanol, the ideal solution is to ingest something that rapidly metabolizes aldehyde. Yet, this second part doesn't yet exist.
There's not a lot of study out there on this subject, and I don't know if any big pharmaceutical companies are working on a aldehyde metabolism drug. But I think the potential market for this is huge. This is one lifestyle drug that I would buy.
I'm currently signed up for a graduate level synthetic biology course here at Berkeley. This is course is cross listed in both Chemical Engineering and Bioengineering departments, and is taught by Professor Jay Keasling.
Apparently, Berkeley has the world's first Synthetic Biology department, and Jay, the founding director, won Discover Magazine's scientist of the year in 2006. Synthetic biology is a neologism for the principled design of molecular biology systems to accomplish a certain task. It is a form of bioengineering (bioengineering is a broader term as it also applies to the fields of tissue engineering or biomedical device engineering). Jay described the discipline with a metaphor, comparing it with building a computer. To make a cell do what we want requires creating millions of different mutants and selecting from these mutants a few that might achieve our purposes. However, from an engineering approach, we should have off of the shelf components that could be assembled into complete biological systems, much like we do with a computer. This is basically the goal of synthetic biology.
The class sounds quite interesting. Most of the people there are bioengineering grad students. We'll have a bunch of guest lectures, some paper discussions, and at the end a group project. Now that the first week of class is over, I'm going to decide which classes to keep and which classes to let go. I think this is one that I'm going to keep (even though it is on Fridays, and we all hate Friday classes).
Today I went to a lecture given by Patrick Moore at the Nuclear Engineering Department's weekly seminar. Patrick was a co-founder of Greenpeace but left because he became disillusioned with them. In the lecture, he strongly advocated a combination of using renewable energy (wind, solar, geothermal) with nuclear energy as the only way to remove fossil fuels from our energy repertoire. And I agree with him 100%. He is an advocate of sustainablility, which mean that we should continue to provide for our needs at the same time reducing damage to the environment (in this case, CO2 emissions).
Patrick, early in his presentation, lamented that the environmental movement has been taken over by extremists (what I call environmental fascists). He summarized these people as: "anti-human, anti-science and technology, anti-trade and globalization, anti-business, and in general, anti-civilization." The story he told was that he once saw a protester with a sign that said, "join the global crusade against globalization" (the joke is obvious... but apparently that protester never got it). These neo-environmentalists aren't really pushing an environmental agenda, but pushing a neo-marxist or even a neo-anarchist one disguised as an environmental agenda. They oppose all things fossil fuel, but also oppose nuclear and hydroelectric energy. Together these consists of 99.2% of the world's energy production. No amount of solar panels can make up for the rest (besides, the production of solar panel themselves causes a lot of pollution).
I have a friend who buys into the whole environmental fascism ideals and it's incredibly frustrating talking to her about this kind of stuff. She believes that nuclear energy is the devil, and that we should go back to living in the 19th century (for some reason she doesn't think that's an extreme position). She thinks that we have disturbed the balance of things (which is probably true), and wants us to go back to what is "natural". But really, what is natural? Living in caves, with an average life expectancy of 15? She believes all life has equal value, which is a fine concept until you really sit down and think about it. Is a human life worth the same as that of a spider? Is it really?
I consider myself an environmentalist. I really believe in reducing greenhouse emissions, clean water, better conservation, and all that. But I care because of a simple (and selfish) reason: if we don't take care the environment, the human species will be wiped out. This is the most logical and most rational reason. I don't care about some random beetle in South America whose name I can't even pronounce. But it is possible that type of beetle (or some type of archaebacteria) holds the key to our survival, or they can tell us something about our place in the world. In that case, their continued existence becomes my concern. It's not that I place the same value on that beetle as I do a human child... absolutely not. Every organism is selfish in that they're only concerned with self-preservation and we should be no different. It's just that since we hold so much more power we need to be intelligent enough to look a few steps beyond the immediate future and do things that don't hurt us in the long run. But this absolutely doesn't mean we must forfeit progress to save the environment.
Nuclear power is in fact extremely safe (it's safer to work in the nuclear industry than it is to work in financial services). This is because safety is built into the nuclear engineering culture and so much engineering has gone into it. Everyone thinks about Chernobyl, but only about 50 people died as an immediate result and only about 9000 are expected to get cancer. Compare this with Bhopal where 3000 died immediately and 15,000 died subsequently. But for some reason nuclear energy has a worse reputation. Chernobyl happened because of bad engineering (the reactor wasn't designed for civilian use) and bad operators (all safety features were disabled for a reactor test). The newest generation of nuclear reactors are nothing like that. With nuclear waste reprocessing, we have basically found a perpetual energy source (nuclear waste isn't waste at all... 95% of the energy remain in them, and with reprocessing they could be put back in the reactor).
I'm most excited about small reactors like the Toshiba 4S. I think we can completely decentralize energy production and eliminate the wasteful practice of transmitting power over long distances. This way, not only will we have more cost effective energy, it will be reliable and fault tolerant. I can't wait to live in that future.
Earlier in the week, I saw this news article saying that scientists in Denmark released the most comprehensive study to date on the link between cancer and cellphone usage. The word on the street (which was confirmed by this research) is that cellphones don't cause cancer (or more precisely, cellphone use and probability of developing cancer are not correlated).
I finally found some time, and read the actual paper that was published in the Journal of the National Cancer Institute. I'm very impressed by it. The researchers tracked 420,095 cellphone users in Denmark and cross-referenced them in the Danish cancer database (which records every case of cancer in Denmark since 1943). The cross-reference is done via a type of national identification number that apparently is assigned to every person living in Denmark. Most importantly, the study follows users over long periods of cellphone use (8.5 years on average, maximum is 21 years). The conclusion is that there is no statistical difference between the cancer rates of cellphone users and that of the rest of the population.
However, like every scientific study, it's not waterproof. Even though I'm very convinced that the methodology and conclusion are both sound, I do like to point out a few areas that skeptics would mostly likely to use to attack this research.
First, although the study looked at over 420,000 cellphone users, most of them were men. In fact, only 62,542 were women (that's more than a 5 to 1 ratio). More over, the cancer rates in women cellphone were slightly higher than average (but not statistically significant). This is explained by the fact that female cellphone users had higher cases of cervical and kidney cancer than the reference population. As unintuitive as it sounds, it may have something to do with the socio-economic status of female cellphone users. The researchers did try to account for socio-economics by looking at income, but there could be other confounding factors that were not taken into account.
Additionally, since cellphones are still very new, only 56,648 users had been using it for more than 10 years. The research showed that there is no increased cancer rates even among these long term users. However, this sentence jumped out at me when I read the paper:
... risk among long-term users could not be completely ruled out because a statistically significant risk (OR=1.8) was found for tumors that developed on the same side of the head as reported the phone use.
As long as our knowledge of cancer is incomplete, we can never know for certain. No scientific study, however rigorously conducted, can completely convince everyone. So I expect the love affair between cellphones and cancer to go on for a long time to come.
And speaking of things I should've learned but never gotten around to, last week I finally learned just what the hell ANOVA is. In all my statistics classes from high school through college and into grad school, ANOVA (analysis of variance) has always been skipped for some reason. Last week I was reading a biology paper (for fun) and they used ANOVA, so I finally hit the books and learned what it's all about.
Simply put, ANOVA is a mechanism to test for correlation in an experiment where the independent variable (the knob that is turned) happens to take on discrete values. If the independent variable is continuous, we can do something like linear regression and measure correlation by looking at ρ². If the independent variable is discrete, one way we can test for significance of correlation is to do a bunch of pairwise t-tests. However there's a major problem with this approach. For example, if the independent variable take on 6 different possible values, then the hypothesis will look like this:
null hypothesis: μ1 = μ2 = ... = μ6
alternate hypothesis: μ1 ≠ μ2 ≠ .. ≠ μ6
To reject the null hypothesis, we need to perform 15 t-tests and suppose we do this at 95% confidence for each t-test. And further suppose that the null hypothesis is actually true, but we cannot accept it unless all 15 t-tests pass. Since we assume independence, this means that we will have a 1 - (0.95)^15 = 0.537 chance of failing to accept the null hypothesis when it is true, which is much more than the 0.05 error for each individual test. This is obviously no good.
The solution is ANOVA. The idea is actually very simple (at least for the one-way ANOVA). When our variable take on c different values, we can get m X c = n observations. We first calculate the variance within the c treatment levels, this is the between sample variance.
S²B = m ∑i (Yi. - Y..)² / (c - 1)
Next, we find the within sample variance, which is another way to estimate σ²:
S²W = ∑i∑j(Yij - Yi.)² / (n - c)
If the independent variable has no effect upon the dependent variable, we expect these two different estimators for σ² to have a ratio of 1 (ie. they are the same). However if the independent variable actually changes the dependent variable, then we expect this ratio to be greater than 1. This ratio is the F score:
F = S²B / S²W
So now we look it up in a table, which should tell us at which F score given the degrees of freedom c-1 and n-c we should reject or accept the null hypothesis.
Ah, the beginning of October. Yet again, it's Nobel Prize announcement time. I love it how the Nobel Prize committe spread out the news over an entire week, so each category's recipient gets their own special day under the sun.
Yesterday, I was very pleased to hear that the discoverers of RNA interference won the Nobel Prize in Physiology and Medicine. I first learned about RNAi in my upper division molecular biology class, and I thought that the experiment that demonstrated it's existence was rather elegant. RNAi, for those not yet in the know, is a phenomenon where if you introduce a double stranded RNA that contains identical sequence with an actively expressed gene, the dsRNA will interfere with the expression of said gene. The discovery was in fact by accident, and has great implications. Already there are companies trying to use RNAi to cure everything from AIDS to cancer. The significance of this phenomenon is so profound that the Nobel Prize was given only 8 years after its discovery. In contrast, Barbara McClintock had to wait over 40 years for her Nobel Prize.
Today, it was announced that John Mather and George Smoot won the Nobel Prize in Physics for their contribution in confirming theories about the Big Bang. A lot of people at Berkeley got really excited since both recieved their PhD here, and George Smoot is also a physics professor here at Berkeley. In fact, he is the 20th faculty of Berkeley to receive the Nobel Prize. He is now eligible to recieve his very own "NL" parking permit (Berkeley reserves the best parking on campus for "NL", ie. Nobel Laureate). So that's pretty cool.
I still remember the first time I heard about the Nobel Prize. I was three years old, talking a walk with my grandpa, and he told me that the highest award for scientists in the world is the Nobel Prize. I don't remember much from when I was three, but for some reason, I remember that conversation vividly. Ever since, I've had a fascination with all things Nobel Prize related. Is it possible that I'll have my chance one day? Hey, I'm still young... and I'm allowed to dream.
