Who moved my heritability, and why should you care?

If I may be allowed to riff on Spencer Johnson’s very interesting title: ‘Who Moved My Cheese?’  let’s have a go at teasing apart a phenomenon gathering some serious interest among geneticists, and others concerned with heritability and why it so often seems to have gone missing.

Missing heritability has stormed to the front of scientific consciousness in only the last several years.  If you type ‘missing heritability’ into Google Scholar you find a 2009 article in Nature by Manoilo and others [Finding the missing heritability of complex diseases] that as of this writing has been cited by others 3423 times.  This is an amazing metric.  The next most cited article in the search I just did gathered a mere 894 citations.  It’s not likely another article will take the top spot for that particular search phrase.

Let’s get on with wondering how all the heritability got misplaced in the first place, and consider just how much effort should be spent building a search party to go looking for it.  First, consider a definition of heritability (not trying to insult anyone’s intelligence here, but this concept gets pretty beat up on a regular basis so I want to give everyone a fair starting point):

Heritability is the proportion of observed differences on a trait among individuals of a population that are due to genetic differences. Factors including genetics, environment and random chance can all contribute to the variation between individuals in their observable characteristics (in their “phenotypes”).

Values for heritability range from 0 to 1.  If a trait has zero heritability it has no relationship with genes, and if heritability is 1 then not only are genes involved, but genes account for all observed variation.   And I might add that if heritability is 1 then it’s likely there is only one gene (not genes plural) involved because when more than one gene is involved there will likely be interactions and heritability will slip.

Why care about this heritability – after all it only goes from 0 to 1 – how boring is that?  Well, some of us spend a good deal of time trying to make better plants and animals to provide food and other products that can feed mankind.  And others among us spend a good deal of time trying to decipher how disease works.  Understanding how much heritability is at play influences our progress.  It is important.  If it just up and walks away we need to get a grip on where it went and how it got away in the first place.  If only H&R Block were advertising:  Get your heritability back America!

So this heritability thing has been around for a good long time.  How is it we’re only just now finding some of it has gone missing?  Genome wide association study (GWAS).  This is a mouth full.  Simply put, GWAS is a technique to take a whole bunch of individuals from some population (a bunch of fruit flies, humans, or maybe even soybeans) and measure them for some phenotype and simultaneously measure their genome sequences.  Then one simply lines up the gene sequence against the phenotypic values, does a regression analysis, and uses the regression coefficient (which also only ranges from 0 to 1… nice how that works 🙂 ) as an estimate of heritability.  What could possibly go wrong?

The whole idea is really not that contrived.  But there are a bunch of moving parts, and as smart as we might like to believe we are there is still more to getting from genotype to phenotype than GWAS will tell us at this point in time.  It is complicated.  But it hasn’t been a wasted effort.  In fact if you consider all these ‘moving parts’ and how enormous the datasets are it might be reasonably concluded it was a miracle we didn’t lose all the heritability!

But the fact remains, we are still looking for a lot of missing heritability.  If you find any, please send me a note.  I’m sure those who need it most will appreciate your effort!

Next up – epistasis in the library with a knife!  (a Clue?)



  1. I’m apparently missing the cognition switch to follow this one. However, it did occur to me dear budding and prolific blogger that you should select the plugin for a menu listing of your postings. You can set it for any number of posts, 5-10 being typical. You might also select the calendar function. That gives your reader a sense of how often to expect a post. Finally there is a number of plug-ins that gives you a global indicator of your audience, as it builds.
    Glad to see it off the ground.


    1. Ooops. And I really dropped the ball in the middle. I should point out there are many characteristics for which heritability is fairly well understood. Take heights of humans for example. We have a fairly good handle on how well height is modulated by genes. But when GWAS methods are applied many studies come up with heritability estimates that are lower, or account for less of the heritability than we’d expect to find (thus the ‘missing’ aspect). Progress has been made finding some of the missing – but that can be left for another time.

      And even this small addition doesn’t do much to better lay a foundation for diving into the subject in the first place. So I’ve left quite a bit of room for improvement.

      Thanks for the additional plugin suggestions.


  2. […] is a picture more easily painted than say holding forth on the merits of various plot protocols for heritability estimation.  With this picture now in place some might smile and change the subject.  But there are still […]


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: