Skip to content

Out of your mind


A local newspaper recently got pretty excited by a press release from my alma mater. It seems a Newcastle bioinformatics team has crash tackled the numbers and come up with 108 genes statistically linked to schizophrenia. That beats the old score by more than seventy. At this rate it shouldn’t be long until they’ve linked all our genes to mental illness.

Team leader and former anti gene patenting activist, Professor Rodney Scott, was quick to clarify that the genes he’d found can’t be used to predict or confirm schizophrenia. Rather they emerge as correlations from large DNA databases of both schizophrenic and ‘normal’ people.

The genes don’t predict the diagnosis, the diagnosis predicts the genes. Sort of. For some people. If there’s environmental triggers too. Maybe. Or it might be multiple sclerosis instead.

But the genes are associated with neurotransmitters, so they probably effect the brain. And schizophrenia is a brain disease. Or an autoimmune disease. Or due to infection. Or drug taking. Or stress. Or a bad childhood. Or poor nutrition. Or the wrong attitude. Or karma. Or something.

It can be pretty difficult doing research on mental illness you know. One problem is that there is no test you can do on your sample population to definitively confirm or deny depression, schizophrenia, bipolar or any other mental illness. No blood test. No urine. No DNA. No biopsy. No neurology, immunology, bacteriology or virology. No scans. Not even psychedelic moving pictures of brain functions.

Mental illnesses are not descriptions of an identifiable physical pathology. They are clusters of behavioural symptoms from the Diagnostic and Statistical Manual. They are what you say and what you do and and what you perceive and how you think.

They are your mind.

But what is your mind?

Is it emergent phenomena of your brain?
The chemicals in your head?
Is it the inevitable product of your genes?
Is it what you eat or what drugs you take?
What you’ve experienced?
Have learned? Have done? Has been done to you?
Who you love and what you hate?
The causes and effects?
All that leads to it and all that comes from it?

Is it what you believe?
Is it only what you believe?

Maybe it’s all those things and more.
Maybe it’s everything that ever was or will be.

Maybe …
your mind is everything …
and everything is your mind.

See. I told you it was all crazy.

From → confusion, DSM, mysticism

  1. Hi. Nice post. I work in the brain ischemia field and just did a study looking at gene expression. Perhaps you know, the whole exercise is based on setting arbitrary statistical cutoffs as to what is and what is not statistically significant.

    The basics of how these things work is fairly simple. One looks at the level of gene x in the normal condition (x1) and again in the disease condition (x2). If the ratio of x1/x2 = 1, then the gene didn’t change. The trick is getting the ratio reliably.

    Experimentally, one will do replicates of each condition. Then, there are a number of statistical methods to determine the ratios. Each has strengths and weaknesses, but with robust results (ratio >> 1) they all pretty much converge.

    But with any use of stats, you must set your threshold for what is significant. Often in biology, it is set at 5% (the famous p 1 at the indicated statistical cutoff. Therefore you can see, it is quite arbitrary what constitutes a statistically significant gene.

    What makes our study unique is we measured >>every<>network<<. The 41 genes are hub nodes and the remainder are part of lesser branches of the network.

    The overall conclusion is that one cannot point to some arbitrary subset of genes and say this "causes" or even "correlates" with the condition. They form a network where the contribution to the network approaches the limit of (ratio = 1) for perhaps half the genes in the organism, and the other half are inputting some nontrivial contribution.

    If this is true for an acute brain disease like ischemia, it is certainly true for a chronic condition like schizophrenia.

    Also, liked your questions about what is mind.




    • Hi Don,

      I approved this comment and trashed the other two pending your fix to the problem of posting tables. If you make it work I’ll be copying from you. I’ve never been able to imbed graphics in WordPress comments.

      Thank you for the kind words.

      I don’t think I follow what makes 41 genes ‘hub nodes’, nor even what defines a ‘network’ of genes. Perhaps you could suggest a link that might help me. Be gentle. Other than 1970s vintage biochemistry my only experience in genetics is the application of population genetics to forensic DNA analysis (Don’t blame me that the field is so rubbish. As with much applied science, politics usually trumps scholarship).

      Often in biology, it is set at 5% (the famous p 1) at the indicated statistical cutoff. Therefore you can see, it is quite arbitrary what constitutes a statistically significant gene.

      Worse than arbitrary when you’re crunching thousands or millions of pairwise comparisons between genes and pathologies. ‘Significance’ then becomes very likely to be coincidence unless you set the cutoff so low you get nothing at all. That’s how deCode managed to find genes for just about everything they checked against their dB of the Icelandic population and why pretty much none of their ‘discoveries’ panned out. Mind you, it did allow them to get in first with numerous gene patents and I suspect that was quite a money earner for them. Until they went bust.

      If this is true for an acute brain disease like ischemia, it is certainly true for a chronic condition like schizophrenia.

      Actually I’d have guessed that linking genes to cerebral ischemia would have shared at least one almost insurmountable problem with linking them to schizophrenia. There are lots of things that can cause it and not all have a genetic basis. Heck, even getting lynched can lead to ischemia and I’m not aware of a genetic basis for that, despite African American heritage being a risk factor as per sickle-cell anemia linked ischemia. Damned tricky that correlation vs causality stuff.


      • Hi (sorry, I didn’t get your name)

        Thanks for emailing. I’ve attached the original post. As you can see, there weren’t even any tables. I just made seperate lines with the mock tables. Its all just plain text! I have no idea why the comment box wouldn’t take it. If you can post the attached and delete the one you approved (or let me know what to do!), I would appreciate it.

        As to your question about networks, I’ve attached a paper of one of my colleagues, Sui Huang. Its surprisingly easy to read as Sui is a excellent writer. Right now, at this moment, there is a slow transition occurring in molecular biology/biochemistry where the idea of “pathway” is being replaced by the idea of “network”. It is a serious improvement as the pathway thinking has become a Tower of Babel that makes no sense.

        A “hub” is a feature of a network. Networks are very easy. There are two components: nodes and links. Links connect the nodes. A hub is a node with many links (as opposed to a node with only a few links). A pathway is an example of a network. The arrows one draws between molecules are the links and the molecules are the nodes. The difference is that there is a whole mathematics of networks that can be brought to bear. This is very powerful and can provide much deeper insight than just listing a qualitative pathway.

        Just for the heck of it, I’ve also attached a paper we have in press that discusses these issues, mainly the limits of the pathway view of things, and how the network math gives a way better handle on understanding. The attached is a “galley proof”, a copy of the article before publication that we edit for mistakes, so please do not distributed it.

        Re the statistics. I’ve been learning about this. The invention of microarrays (gene chips) has necessitated the development of new statistical calculations for analyzing these chips. Like anything, they have strengths and weaknesses. In general, they work better than the old pairwise methods, but are by no means perfect.

        I take your point on causes of brain ischemia. The thing is, after it has occured, no matter what caused it, the response of the brain cells is quite stereotyped. There are of course variations, but they are variations on general themes. Furthermore, the stereotyped responses (called, in general, “stress responses”) are common to a variety of brain pathologies. There is a limited repertoire of stress responses coded in the genes of the brain cells. Depending on the disease state, the different stress responses will be expressed in different proportions, but the players generally are the same.

        Now, it is a very serious issue what is the relationship between acute injuries like ischemia, or traumatic brain injury, and chronic injuries like schizophrenia (or Parkinson’s or Alzheimer’s). You will notice in my writings I am always careful to specify “acute injury” because it is still unclear how acute and chronic injuries do or do not overlap in terms of brain cell responses.

        Well, very nice to meet you and thank you for taking an interest in my comment. I hope we can work the bugs out of it. Please reply only at your convenience.

        Best wishes,



Over to you

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: