This 2005 review article notes that
Over the past 15 years, more than 2,000 quantitative trait loci (QTLs) have been identified in crosses between inbred strains of mice and rats, but less than 1% have been characterized at a molecular level.
I wonder if anyone knows of an updated reference, or a way of updating that estimate. Thoughts?
To my knowledge, there has not been a survey published since then (e.g. in NR: Genetics, Mamm. Genome, Genetics, etc.) that directly updates this question. I think the fundamental point, that only a small proportion of the genetic effect behind mouse QTLs identified using genetic crosses have been convincingly tracked to a specific genetic variation, still stands. There have been some excellent papers since then using yeast and arabidopsis (e.g. Kruglyak's work in yeast, Jiménez-Gómez et al PLos Genet 2010, Gerke Science 2009) where the powerful genetic tools available in those models allowed the groups to identify and test candidates. It's still a slow business with mice, especially since it's reasonable to argue that disease traits (particularly those affecting cancer susceptibility) are more time-consuming (and, possibly, more complex) to dissect than the genetic traits affecting normal physiology that are often the subject of those studies.
Practically speaking, to survey existing QTLs you can head over to JAX's informatics site and pull down data. The set of those which have been successfully characterized is more work and, I think, partly a matter of opinion.
David is pointing you in the right direction here. Let me add a couple specifics.
One, some groups are looking at the intersection of QTL data for clinical-type phenotypes with gene expression data. Or, which genes mapping in/near the QTL peak show differential expression (under certain conditions). In this regard, I would look at the papers from Bev Paigen (Jackson Lab) over the last few years. She has used this approach to get the gene or a region of 3 genes for a number of QTLs pertaining to blood lipids, but, as David writes, it is slow and not always apparent which gene and which variant is responsible for the phenotypic difference driving the QTL. Some of Paigen's QTLs have not yielded to this narrowing approach.
Another option is to use synteny to refine the QTL. I write a lot about synteny here because I feel it is overlooked but is a great advantage to employ when you have completed or nearly completed genomes at hand. I have compared human, rat and mouse (sometimes pig) QTLs for the same or very nearly identical traits as a way to trim the QTL to something smaller in physical distance. I do realize that this is not getting you to that single gene as in it that older review article.
Larry, thanks for response. I have a follow-up question (that is a bit tangential to both my original question and your answer). When looking at synteny, what QTL databases do you consult? RGD is very straightforward, but MGI can be opaque for me at times to download QTL descriptions. And human QTL?
Andrew, I agree with your assessments. I was (still am) interested in specific phenotypes. I bit the bullet and searched MGI manually by keyword. This was right around the time the mouse genome was announced and so choices were limited. On the human side, I found a database specific to our interests and also used tools/data at NCBI. THe latter seem even better now and I'd use those for any updating I would do. My goal was to obtain the boundaries of the QTL in terms of genes (not bp coords). If you have more questions, send me an email.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
great articles and links, and more importantly, great perspective from someone in the field... thanks!