Wherein I #FAIL personal genomics….
One of the medical issues of personal interest to me is allergy. It’s something I pay attention to all the time because I have an allergy that can really cause me problems–peanuts. But I come from a family with a whole host of additional allergy issues too. All of my siblings have had serious problems with eczema. Their cracked and raw winter hands can be significantly painful, and I feel for them. Sometimes I felt like the peanut allergy was a better bargain–not having a Reese’s peanut butter cup is just not that big a deal, ya know? Luckily I haven’t been fooled by any peanut flour that was able to kill me–so far.
So I when I read the recent article in The Scientist on The Allergy Gene I was totally engrossed. It’s not your typical research article–it’s the story of the work and the outcome in a narrative way. It’s about how this group went after an incredibly challenging gene–filaggrin (FLG). It is a cytoskeletal protein–an intermediate filament associated class one. As my PhD project was on cytoskeleton I’ve also had a professional interest in these too. It’s also a massive mRNA, making a “monster” pre-protein–making it rather unusual. And nearly the whole thing is coded by a single exon. That’s even stranger. There’s also some really heinous repetitive stretches–a sequencing nightmare. So from a genomics/sequencing/gene structure perspective–wow! That’s a cool story.
But it makes sense to me that if the family epidermis or other surfaces are a bit more porous that some other people’s, that could explain our contact allergy issues.
So FLG is compelling, curious, and intriguing to me from a number of perspectives. I was planning to write this up anyway because I thought people I know would be interested in the allergy gene–when just last week I saw another fascinating piece come along, from the BBC of all places: Scientists claim peanut allergy ‘gene flaw’ link. Guess which protein it is? FLG. Bam!
Like most gene-link stories, this will only be part of the explanation. But it looks like it could represent a significant number of the cases of peanut allergy with a filaggrin defect.
I began to wonder to myself: what does my filaggrin gene look like? I could look it up using the 23andme data. But what SNPs will that have? If this is a nasty and repetitive region, how good are these SNPs? And what are the key pieces of FLG that I need to know about?
So I turned to the 23andme data set, looking for FLG. I get a list of SNPs. Some of them have dbSNP ids. Some of them don’t. Ok… [PS: I'm not disclosing my personal SNP data here as I don't think the security and protection of my information is sufficient; I have disclosed my peanut allergy because it is already part of my medical record anyway and can't hurt me any more. ]
Now what? I have to go look at the positions specifically in the UCSC Genome Browser to get a better handle on this, especially for the non-dbSNP ones. Ok–I write up a quick custom track with my data (which I’m not showing you).
Whoa–look at that odd FLG gene region in the UCSC Genome Browser (load up my UCSC Genome Browser session showing this region by clicking here). There’s that massive exon, and check out the conservation track. That is one of the strangest looking gene situations I’ve seen–and I’ve seen a lot of genes. Bizarre.
It’s loaded with SNPs–many coding non-synonymous changes. If you open the SNPs 131 to “pack” mode, you’ll find one of the biggest SNPs I’ve ever seen. Look at the red box for rs 71770072 . It’s a “large deletion” of almost 1kb (but PS: I have seen larger; try in hg18/SNP130 rs 71582232; and why do my snp ID keep vanishing when I edit this post???). Some of the SNPs match the 23andme data. Most don’t. There are just a couple of handfuls of 23andme data.
But fine–I’m looking around now for other tidbits of information. Let’s see what we can learn about the mutations from the FLG paper. I check out the peanut allergy paper, which is fine. But they refer to previously characterized mutations, so I have to hunt back some more. On my hunting expedition some papers are behind pricey firewalls, so I don’t go for those. But I found one paper in The Journal of Investigative Dermatology that had a summary diagram of the mutations. Psyched!
I check out the figure with the mutations illustrated. Ah, damn. They are all cryptic names–protein position related, not handy dbSNP ids. No idea what assembly/version of this protein sequence. Sigh. So I start tracking them down. I’m checking OMIM. EntrezGene. Leiden Open Variation Database. HGVS. SNPedia. Psyche! Nothing really useful here. Sigh.
I’m coming up with nothing that helps me to understand the relationship of the new data, the old data, and my personal genomics data. And I know what to look for. It’s been a pretty frustrating few days. And I have a flashlight. Most people who open up their personal genomics data right now are in the dark, without a flashlight. We’re so not ready.
If anyone else takes a look a their FLG and figures out what mutations match the clinical data, let me know. I’m going to keep looking, but I have other stuff I need to do. I don’t have time for this, for one gene, for now. Sigh.
I #FAIL personal genomics. I have no idea what my filaggrin gene alleles correspond to. That’s embarrassing.
Sandilands, A., Smith, F., Irvine, A., & McLean, W. (2007). Filaggrin’s Fuller Figure: A Glimpse into the Genetic Architecture of Atopic Dermatitis Journal of Investigative Dermatology, 127 (6), 1282-1284 DOI: 10.1038/sj.jid.5700876
Brown, S., Asai, Y., Cordell, H., Campbell, L., Zhao, Y., Liao, H., Northstone, K., Henderson, J., Alizadehfar, R., Ben-Shoshan, M., Morgan, K., Roberts, G., Masthoff, LJN., Pasmans, SGNA., van den Akker, PC., Wijmenga, C., Hourihane, J. O’B., Palmer, CNA., Lack, G., Clarke, A., Hull, PR., Irvine, AD., & McLean, WHI. (2011). Loss-of-function variants in the filaggrin gene are a significant risk factor for peanut allergy Journal of Allergy and Clinical Immunology, 127 (3), 661-667 DOI: 10.1016/j.jaci.2011.01.031