Uncommon Descent Serving The Intelligent Design Community

Ginning up dangers of Neanderthal genes

Denyse O'Leary
February 26, 2016
Human evolution, News
Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email
File:Homo neanderthalensis adult male - head model - Smithsonian Museum of Natural History - 2012-05-17.jpg
Neanderthal 70-80 kya/Tim Evanson, John Gurche

From New Scientist:

Tony Capra at Vanderbilt University in Nashville, Tennessee, and his colleagues studied an anonymised database containing medical records and genetic data from more than 28,000 people of European descent. Europeans retain some Neanderthal DNA, though the amount and the exact parts they have vary between individuals.

They used information from the Neanderthal genome to identify segments of Neanderthal DNA in each person’s genetic data. Then they explored whether particular chunks of Neanderthal DNA in their genomes were associated with a variety of specific medical conditions.

They found that certain Neanderthal segments come with a small but significant health risk: people carrying them are about 2 per cent more likely to develop depression and 1.4 per cent likely to have a heart attack than people who lack them.

More.

But are 2% and 1.4% statistically significant increased risks? Wouldn’t we expect at least around 5%, to notice? What are we missing here?

As well, a large proportion of people with a higher proportion of Neanderthal ancestry live in northern environments where, say, alcoholism and depression correlate reliably with long, dreary winters.

Here’s the study.

See also: Neanderthal Man: The long-lost relative turns up again, this time with documents

and

A deep and abiding need for Neanderthals to be stupid. Why?

Follow UD News at Twitter!

Comments
No. Any change in odds, even 0.000001, may be significant. The typical value (leaving the controversy on significance levels aside) , alpha = 0.05, is related with the confidence intervals. So, the point estimate in increase in odds is 0.02 (2%), and if it's deemed significant, it's 95% confidence interval may be something like between 0.005 and 0.030, not including 0. I guess this is logistic regression, hence the upper and lower bounds are not symmetric. However, this is also an observational study, hence assigning causality is also pretty meaningless, especially at such a low effect size. There may be many unaccounted factors which may "cause" depression. That's why it reads in the link: the research has identified correlations between Neanderthal DNA and medical conditions rather than demonstrating causative links. Those correlations are very robust, he says, but they don’t prove that the Neanderthal DNA itself is directly responsible for the heightened risk of disease. The word "prove" (Popper must have turned in his grave) is wrong again, it should be replaced with "show". Well, there's correlation, there's no causation, but the whole news and its relation to Evolution is based on the idea of causation. That's a very good example of today's science.. What is implied is much much larger than what can be rigorously shown... CuriousCat
Haha yes. The 2% is effect size. Presumably some other significance measure secures this. Meaningless study though, I would expect. Splatter
#2 wd400
The slightest bit of knowledge about how statistical inference works.
First big smile of my day! ellazimm
But are 2% and 1.4% statistically significant increased risks? Wouldn’t we expect at least around 5%, to notice? What are we missing here?
The slightest bit of knowledge about how statistical inference works. wd400
what we miss is the usual... we make the research interesting for medical purposes...so we increase the chance of funding. This is normal in the post-science age. Peer

Leave a Reply