week_of_science: genetic linkage and psychosis

People went, on the whole, crazy. The riots and rallies in my city resulted in buses burnt, people hurt, the usual sad story. I wondered about the genetic basis of psychosis. As a matter of curiosity , I decided to look into the genetic basis of insanity. What I found was interesting insights into a couple of diseases that I am very interested in.

Everybody knows that DNA is the blueprint of life. It contains, coded as a sequence of chemical elements, all the information that one needs for life. It would not be straying TOO far from the truth to say that genes – sections of DNA – determine most of our characteristics. At least, predisposition towards certain characteristics.

It is well known that the genes linked to insanity (which by itself is a broad term, and cannot simply be used as I have here) have not been identified. In fact, there are so many implicated (alongside a number of other variables such as environmental factors) that it is generally accepted that a simple solution does not exist. I ‘know’ this, having somehow absorbed it by university-atmosphere-osmosis. The question I asked myself, therefore, was: is there some sort of proof of this complexity?

You could see disease as a state of abnormality. This abnormality comes, presumably, from changes in the DNA – small variations or polymorphisms that code for, say, the natural colour of your hair. It is generally accepted that DNA variation is due to the effect of one or more genes. It is relatively simple to detect a single gene that controls a feature. You just scan the genome of a (large) bunch of people who either have the trait or don’t. If you see a variation that is consistent in all of them: for example, if gene X is expressed in all individuals with condition Y, and not otherwise, then you can conclude that gene X controls character Y. In the case of multiple genes exerting an effect, the procedure is a bit more complicated, but in effect, you can detect a similar pattern. Genes X, Y and Z must be active, and gene A and B must not, in order for a condition L to be true. These studies are called genetic linkage studies, and the basic laws of genetics were derived by Mendel using them to study a pea plant. Luckily for him, there were no environmental factors controlling the characters he was studying. This process is infinitely more complicated when you are studying predisposition towards a disease rather than the occurrence of the disease itself, because if a person without the disease has a predisposition, how do you know? You study his family, that’s how.

You could take a pair of identical twins, who have the same DNA and hence exactly the same predisposition. If one carries the disease and the other does not, then you could conclude that there are some external factors affecting the disease. A trigger, perhaps. Timothy J. Crow in his paper* asserts that changes underlying psychosis are epigenetic : which basically means that DNA sequence does not change, but something in the whole big genetic machinery does. Where does one see genetic machinery? In gene regulation – for example, although all the cells in your body contain the genes to produce insulin, only the Islets of Langerhaans do. It usually refers to the modification of the DNA (methylation, acetylation of histones etc) without changing the sequence of DNA.

Crow’s argument is that several large scale studies have come up with rather discordant results: while one predicts one set of gene loci associated with schizophrenia, another comes up with a different set. A study of genetic linkages over a larger dataset actually weakens, and in a particular case, refutes, the results obtained by a smaller study. Basically, no one has been able to come up with a consistent set of genes that could code for the predisposition towards schizophrenia or bipolar disorder. The experiment when repeated on different scales wind up with different results. So, obviously, these are not valid. Given that there is no conclusive evidence for a multi -gene system, his hypothesis is that there is a single gene or set of genes that code for this disorder. Further, since the X chromosome is not considered in any of the genetic linkage studies, he figures that the genes might be present on both the sex chromosomes. He reasons that there is a strong connection between sex and onset of disease, and thus, if the variation is epigenetic, it is likely to occur during meiosis.His theory is that the disease-state is somehow epigenetically triggered, which is why it is not detected by genetic linkage studies.

To me, his most startling argument is the connection of language and the disease. Given that identical twins have a 40-50% chance of having the same disease state, there must be some other factor that accounts for the difference. He notices that monozygotic or identical twins have different handedness – one left handed and the other right handed and that this asymmetric genetic bias is due to the development of celebral dominance. This is also essential for the development of language. It is known that primates are not left-or-right handed (called directional asymmetry), nor, he says, do they have the faculty of language. He further claims that the symptoms of psychosis are linked to language, be it hallucination, disturbance of thought processes or delusion.Epigenetic factors contribute powerfully for individual differences in the asymmetrical development of the human brain (unique to us). Thus, he links language,epigenetic control, and predisposition to psychosis.

He concludes that that the variation that codes for psychosis is actually on the sex chromosome, is epigenetically controlled and is related to hemispheric dominance for language. I, for one, cannot see how he arrives at that last without first confirming that other primates do not suffer from schizophrenia. Because he shows no proof outside logic, and we know that the communicative property (A give B, B gives C, implies A gives C) is not true in many biological situations.

* Timothy J Crow 2007 How and Why Genetic Linkage Has Not Solved the Problem of Psychois: Review and Hypothesis; Am J. Psychiatry 164:1

Disclaimer: This is neither a thorough review of the subject, nor have I covered all the aspects involved in such a subject. I’m only trying to demonstrate how far we have come, and how much further we have to go. I am not an expert in this field, and am only going to explain some basic concepts in this post.

week_of_science: on laughter

One week of writing a post a day about science shouldn’t be difficult, correct? To not fill it with musings that go absolutely nowhere … well, that should be a peice of cake, correct? Well, I signed up for a week of science challenge, with a few brave people – it begins today. And then spent the last day looking up something to write about. I had only this as the guideline:

Bloggers who self-identify as scientists and science writers should post on:

1. Published, peer-reviewed research and their own research.
2. Their expert opinion on actual scientific debates – think review articles.
3. Descriptions of natural phenomena (e.g., why slugs dissolve when you put salt on them, or what causes sun flares; scientific knowledge that has reached the level of fact)

Bloggers who claim to be philosophers of science (or have been accused of so much) should post on issues, ideas, and debates in philosophy of science that are not frequently used or dictated by anti-scientific groups. The demarcation problem, for example, should be avoided unless it can be discussed without reference to anti-science movements.

And bloggers who are not scientists – focusing mainly on public and policy debates on scientific issues – should post on issues that are legitimately controversial for scientific reasons. Topics that are controversial simply because of anti-science movements should be avoided.

More here.

I’m hoping to talk about laughter. And yes, first post about week_of_science in a few hours. Meanwhile, the feed is here:

http://www.justscience.net/?feed=rss2

Lets get started!

Perspectives On Bioinformatics

Despite having spent (almost) four years of my life studying Bioinformatics, I couldn’t precisely define it now if you asked me to. Yes, I can expand the name – the application of computer science to biological question. So what, then, is Computational Biology?*
Stumped.

Alright then, one can say, Bioinformatics is is application of biology to computer science. Like … (grappling desperately for buzz words) … genetic algorithms! And … Um*…
Stumped.

So what’s going on? Do I know nothing? Or do I pull the standard “Bioinformatics is like Life and Beauty. Definition varies from person to person.” (It does, by the way.)
Too corny.

Wait, lets start over, with the standard lines everybody loves to use, “With the exponential increase in the amount of biological data available today, it is increasingly important to use computational apporaches to arrive at solutions to critical biological questions. Bioinformatics is the name of the field that intgrates computer science and biology. It is a multi-disciplinary field, with applications in structural biology, biophysics, pharmaceutical industries, genomics, micrbiology, microarray technology,” and a whole lot more. There is Big Money in it.

Obviously, the result is a disaster, with Universities jumping into “hot”, “cutting edge” courses feet first, some of them (like mine) with all the requisite hardware, but no experienced or established people. No real professors who do real research work. Of course, most universities I know of don’t give profs/lecturers time to do anything but speak, but that is an issue for a different rant. The students who are produced are confused more often than not, and since they joined such a broad based course, have no idea where they are going.

So, with your fingers stuck in several pies, what are the odds that you’ll actually know something in the field? Wouldn’t it be better to do a course in say, biophysics, and then go to bioinformatics? At least in India? (But where is a good biophysics course? And this assumes that you know what you want to do at 17-18. Most of us are not that lucky.)

Bioinformatics, I believe, is not a specialization like most (respected) industry people and research scientists believe. It is a base course to leap into an interesting field, it allows you to have a good general idea, the big picture. It is an amalgamation of a number of streams, and so being rather new, it’s still in the turbulent phase. It is what will one day, I imagine, be essential to most biology related fields, maybe even others, and be integrated so throughly that it might lose its identity as a field. Maybe. I imagine it will eventually stop being an imperfect patchwork quilt of courses and start being a solid base course. It’s like doing a B. Sc. physics when you are confused. You can go where you want. And you have time to figure it out. Four whole years. (Far be it from me to compare it to Physics. Physics is the mother of all science. B. Sc. Biology is irrelevent with the course contents these days. B. Sc. Maths rarely has the type of teachers that one need to be really interested in it. Unless you are in IIT. Then, of course, you are in IIT).

I still think it is an amazing place to be. Only, with industry deciding that B.Tech. students don’t know anything (can’t blame them), and research groups asking, “but what do you KNOW?” (with no chance of launching into the What is Knowledge? and the Transcience of IT All), one can feel incredibly lost for options.

I can’t wait for it to grow up and become a solid stream, like computer science did, so I can say I told you so.

* I do actually know the answers to that. But the average B.Tech. or M.Sc. Bioinformatics student does not.