Tuesday, February 2, 2016

We're all fundamentalists now

If you're a foodie at all, you've heard of Yotam Ottolenghi, chef, restauranteur, and food writer.  Perhaps you've used some of his recipes, or even have one or more of his cookbooks. And, if you're a fan you'll be happy to know that the Jan 11 episode of the Food Programme on BBC Radio 4 has an interview with him (starting at minute 7:15), and a brief overview of how he got to such a place of prominence in the food world.  

Ottolenghi and his business partner, Sami Tamimi both come from Jerusalem, but Ottolenghi from the Jewish west side and Tamimi from the Arab west.  They both now live in London where they have collaborated since the late 1990's on restaurants and delis and cookbooks, much of it with the aim of highlighting the food of their childhood.  Not only is their food amazing, but it's also worth noting that two men from two sides of the same strife-ridden Middle Eastern city have worked closely together for many years. This isn't something that everyone could do.   

One of the cookbooks Ottolenghi and Tamimi wrote together is called Jerusalem, written to accompany a BBC television program of the same name. For the show, they returned to their birthplace and described and prepared some of their favorite foods, but it wasn't just about the food.  Tamimi said it was difficult to return. He believes that people were much more naive when he was a child, having faith that the conflict between Israel and Palestine could be solved. Now, he says, people are much more entrenched in their belief in the rightness of their side, and it's much more difficult to imagine the differing sides agreeing on a solution.

To us in the West, the Middle East epitomizes fundamentalism, strict adherence to the literal interpretation of a religious text or dogma.  And, fundamentalism goes hand in hand with terrorism. Fundamentalism is our enemy.

But, it's not just in the Middle East that people are more entrenched in their beliefs about right and wrong. Here in the US we've got the Tea Party dictating what real conservatism is, we've got militiamen in Oregon, and homegrown 'terrorists' demanding whatever they're demanding. We've got a Congress that agrees only to disagree. Dare I say it, even the 'new atheists' are fundamentalists. Indeed, compromise has become a dirty word, immoral even. In so many ways, moderation, the ability to see more than one side of an issue, has lost its way.

Ken's view is that in a world in which fundamentalists are now our enemy, we've all become fundamentalists; we know what we believe, we hold to those beliefs without question, and we have no respect for the other side.  If there is a strongly ideological force that you disagree with or that threatens you, it pushes you toward an equal and opposite ideology. You listen only to Fox News or MSNBC, you turn off the radio when Trump comes on, or when Clinton comes on, depending on your predilection -- or you're waiting for the Libertarian candidate to be selected, and there's no way you'll listen to anyone else.

If you're here reading this it's likely that you've also picked a side in the nature/nurture 'debate'; 'genetic determinism' either nicely describes your view of biology, or you're very uncomfortable with the term. Genes will or will not be found 'for' most traits, including behaviors, and diseases will or won't be predictable once we've all got our genomes on a CD.

We've said this before but it's worth repeating.  In 1926, one of the great early geneticists, Thomas Hunt Morgan, wrote this about stature:
A man may be tall because he has long legs, or because he has a long body, or both. Some of the genes may affect all parts, but other genes may affect one region more than another. The result is that the genetic situation is complex and, as yet, not unraveled. Added to this is the probability that the environment may also to some extent affect the end-product.
                                  (TH Morgan, The Theory of the Gene, p 294, 1926):
Morgan would be totally comfortable with the recent GWAS results showing that there are hundreds if not thousands of genes that contribute to stature, as well as environmental factors.  He'd agree that complex traits, like stature, or many diseases (including schizophrenia, which Ken will talk about tomorrow) are polygenic, with some environmental effect.  This has been known for almost a century. So why are people still looking for genes (meaning single genes, or a few genes with individually strong effects) 'for' type 2 diabetes, or heart disease, or stature, or schizophrenia?  Why don't we still know what Morgan knew so long ago?


Because sometimes it's not true.  Sometimes there are single genes whose variants are by themselves responsible for traits, including disease.  Starting in the early 1980's, the role of single genes in various traits began to be discovered; oncogenes, Huntington's, cystic fibrosis, breast cancer, and a whole host of single-gene pediatric diseases, and normal traits as well, like blood types, eye color and so on.  There are now about 6000 rare diseases for which genetic causation appears to be known, or at least claimed.   This history of successes mislead, we would say, geneticists, and others, into assuming they could always expect to find 'the' gene for this and 'the' gene for that.  In essence it is still the informal working model, in the back of geneticists' heads, that everything segregates like Mendel's pea traits.

We can and do have both -- single-gene traits and complex traits due to many genes, or many genes and environmental factors too.  Indeed, there are also traits that are completely environmental -- look at the havoc Zika virus seems to be wreaking, with apparently no help from genes, even if close examination might find some people to be slightly more immune than others. Most viruses are like that.

So, it's curious that even the field of genetics has its fundamentalists.  Every time Ken and I write about complexity, or insufficient understanding of disease causation, or question how we know what we think we know, someone will send us a link to a paper that shows we're wrong because autism, or schizophrenia, or intelligence, or whatever their favorite trait, has been shown to be clearly genetic. Genes, with names, have been found to explain it. Sometimes the comments are so emotionally unrestrained that you'll never see them because we don't publish them.

And, we'll often or even typically look at the paper and realize that we've been reprimanded by a fundamentalist yet again. Autism, schizophrenia, heart disease, stature, intelligence, and so on are just not yet predictable from genes, and, we believe, are unlikely ever to be for reasons we write about all the time.  Ken will discuss the new Nature paper on schizophrenia tomorrow, a paper that got huge amounts of press for finally beginning to explain the disease.  Yes, a paper someone offered to send us when they disapproved of a post Ken had written about the difficulties of predicting disease, proving he was wrong. Which, good as that paper may be, is not the case.

I think if Morgan were to come back to the modern field of genetics, he'd feel as Sami Tamimi did returning to Jerusalem.  I think he'd be nostalgic for his era, when fundamentalists didn't rule the field, geneticists weren't prisoners of Mendel, ideologues who know what they'd find before they even looked. Where, even if things seem rosier in retrospect, and certainly people had preferred views and were not always nice to each other, there was more agreement that things were not yet clearly understood, and complexity was not a dirty word.  I think Morgan would appreciate that some traits are explained more simply than others, but that even those aren't 'simple' -- there are more than 2000 alleles in the CFTR gene that seem to be associated with cystic fibrosis, and this kind of complexity is true of most 'simple' traits.

So, why did the field lose this understanding, and take a turn to fundamentalism?  The answer isn't just that we're in a fundamentalist age, of course.  That it's a lot easier to sell the search for a causal gene than a search for.....we're not really sure what, is a large part of the problem.  But, as a friend says, we should be looking for the molasses that explains biological complexity, that connects causal pathways and processes, which ain't just gonna be a gene, or an environmental risk factor.  It's going to be something we don't yet understand, and continuing to look for 'the' gene for your favorite complex trait is only going to slow down the search.  Acknowledging that what we've learned, and confirmed over and over again since Mendel was rediscovered in 1900, is that most traits are complex -- and unpredictable -- is a crucial step.

8 comments:

Bill R said...

It looks like a reaction to me. When I first learned this stuff (the '60s), it was assumed to be complex, with Mendelian cases as special cases, and prediction was hard and inexact. Then the potential genetic predispositions and causes got buried because of the social/political implications of lack of equality. People will still argue against GWAS and studies because of the consequences. The same stuff happens for long term epidemiology/medical studies.

Ken Weiss said...

We do react, because we see fundamentalism and cock-suredness all the time in science. Some of it is just people being people and working their points of view and agendas. But some of it is persistently racist or things similar and societally dangerous. Lots of it is the jostling for funding and other forms of credit. Yes, it's all human, but fundamentalism is usually dangerous or damaging and in science of all places it should be heavily resisted.

Among other things, characteristics of fundamentalism are hubris and exclusiveness. When it comes to public funding, we care a lot, because manoevering for large long-term funding supporting one point of view takes that funding away from others. Well,we say this all the time here, so no need to belabor it in a comment.

This all doesn't mean we can't be confident of our viewpoint and work hard to advance it, because that is a natural way to be. And it doesn't mean we shouldn't strive for funding to help do that. Maybe developing a system less vulnerable to manipulation in this way would be a remedial factor.

Ken Weiss said...

I'll just add that we scientists are well-trained not to over-commit. We know very well how to say, even if buried, that things are more complicated or this or that is as yet unknown. But the assertions are far stronger and more prominent, and the caveats are CYA additions, that always leave us wriggle room in case for some horrendously unexpected reason we turn out to be wrong!

Anonymous said...

"we should be looking for the molasses that explains biological complexity, that connects causal pathways and processes, which ain't just gonna be a gene, or an environmental risk factor. It's going to be something we don't yet understand"

Historically, a lot of scientifically researchers bumping up against a phenomenon, recognizing it, yet coming up with inadequate explanations may be a sign that a major breakthrough - a profound explanation that synthesizes, simplifies, and renders predictive - is forthcoming if not imminent. In this case perhaps it will be some startlingly elegant mathematical wrinkle on complex systems, networks, and/or self-organization. Alternatively, maybe that 'breakthrough' will be a proof that no explanation is possible, that the noise can't be teased apart further; these phenomena will remain forever opaque and the best that can be done is brute computation of associations (much of them spurious).

Ken Weiss said...

This is a thoughtful reply. It's always tempting to be a Kuhnian and say this pressure to get better theory will lead to a breakthrough, a savior in the form of the next Darwin or Einstein. At present, it's not clear to me (as to you, it seems, as well) whether what we see is just what's there, or whether some new basic insight will make a big difference. Are we ignoring of strange results that seem to suggest some form of 'molasses' is yet to be found? I guess until someone finds it, nobody will know.

Lyndon Page said...

To all Mermaid posters, many nice reads lately.

To anon and Ken, I do not get the idea of radical breakthrough or some new fundamental insight. It seems most of the answers are laid on the table. We have a good enough understanding of the basic principles that lead to traits. The complex genetic structures and environment and developmental structures that lead to a given height at a give time for an organism seems like something that we understand from a comprehensive standpoint. The same holds for intelligence, once we slay the mind.

Science may make great strides in complexly detailing those stories. It will not be revolutionary. Unless one is working with some extreme naturist or some extreme nurturist paradigm.

It seems to me, even if something like epigenetics was found to be robust in a surprising way for a trait, it may be revolutionary in our thoughts about DNA determinacy, but it would not necessarily be revolutionary from a biochemical standpoint. That is simply because the complexity of gene expression that we are painstakingly trying to figure out is often separated from the overhead theory.

Ken Weiss said...

Thanks for these thoughtful comments. What counts as 'good enough' is of course in the eye of the beholder. Your ideas certainly seem reasonable. But there are daily raves tempered only with quiet caveats about findings, including promises of 'precision genomic' medicine, with all its usage is intended to imply. We've written about this before. If one simply accepts that, for example, genomes are generally weak predictors for the sorts of traits under discussion and that there is no major concept or functional aspect missing--no 'breakthrough' or revolutionary discovery, then your assessment would be appropriate, but not everyone seems as accepting of that. It would largely imply that we won't or even can't in principle have a highly specific causal understanding or predictability. That's possible and to me, at present, we should acknowledge our current limits and not promise too much, and not keep hammering on the same (costly) drum, and to invest much more resources where real tractable causal (or even 'curative') progress is most likely to be vulnerable to technological approaches; and see what broader discoveries are made in the future, which one can't predict.

Anne Buchanan said...

Thanks for your kind words, and your thoughts, Lyndon. Perhaps it's simply a glass half full/glass half empty distinction? Some very smart people read the new Sekar et al. paper on schizophrenia (see Ken's post today, Feb 3), for example, and see it as a real breakthrough, meaning we'll be able to predict the trait, and with pathways beginning to be elucidated, finally develop targeted medication. Others, like Ken, see that this was very good work, but it tells the same story that almost every other breakthrough about complex traits has told -- the trait is complex; it's polygenic, there are presumably environmental components or triggers, people without the trait can have the 'causal' allele/s, there are many pathways to a single trait, and genes do many things. We can't predict environmental exposures, everyone's are unique anyway, and everyone's genome is unique, and when you throw in possible epigenetic and even microbiomic factors, prediction, or even understanding the cause of any one person's disease -- or why another person doesn't get the disease -- never mind all causes of the disease, begins to seem daunting. And, people talk about gene by environment interaction, but in most cases it's not yet clear what that actually means.

Perhaps complex traits are 'emergent properties', whose causation we won't understand even were we do in fact know all causal factors. And, of course, before we found small RNAs were discovered, or epigenetic modification, or the effect of the microbiome on a variety of traits, and so on, we thought we understood theory well enough, too. These may not be 'revolutionary' findings, fundamentally changing how we understand developmental genetics, but they do add to our understanding in ways that we now can't ignore. Similarly, it's impossible to know if there's some 'molasses,' as our friend says, that will give order to what now looks almost hopelessly complex and unpredictable.