What are the rules about naming genes?
In a recent Naked Genetics podcast Dr Kat talks about genes called "unkempt" and "headcase". This caused me to wonder how they came to have such bizarre names. Then I wondered what the "rules" are for naming genes? Who gets to christen them? Is there a classification system and agreed nomenclature for genes (as there are for plants and animals) or can people who discover them call them whatever they like? Surely a free-for-all would be chaotic - no-one would know which gene was which, there would be some genes that were "discovered" by several different people all of whom called it by a different name, and possibly lots of different genes with the same name (although I doubt whether too many people would name their newly discovered gene "headcase"). Thank, Nicky
Kat - At this point we'd usually be hearing about our gene of the month - over the past couple of years we've had some wonderfully-name examples, from Sonic Hedgehog to Superman. Most of these are named by researchers working with model organisms such as fruit flies, who tend to pick a name inspired by the appearance of a fly lacking that gene. But listener Nicky Peng wants to know more and has written in asking: "Is there a classification system and agreed nomenclature for genes (as there are for plants and animals) or can people who discover them call them whatever they like?"
I spoke to Elspeth Bruford, who leads the HUGO Gene Nomenclature Committee - they're the people who get to decide the names of human genes - to find out what's in a gene's name, and how they pick them.
Elspeth - Obviously, we take into account if a gene has been published. We try to discuss with the authors who have published on that gene. If a gene hasn't been reported in any scientific publications then really, it's up to us and we have to look at features of the gene and decide how we think it should best be named.
Kat - Now, some of the genes that I've featured in the podcast have wonderful names because they were discovered in things like fruit flies. So, you have genes like Eyeless or Wingless and all these kind of things. Those aren't the kind of names that come through as human gene names when the equivalent gene is found in humans, are they?
Elspeth - Sometimes they are but then really, we try not to use the more whimsical names that some fruit fly researchers have chosen to use. Mainly because the aim of human gene symbols is that they should be used in all context. So, not just in scientific publications or presentations or discussions, but also - and you see this more and more - that our gene names should be used in the media, they should be used by clinicians in discussions with patients and by GPs. And especially if you're in a scenario where you're telling a patient or the family of a patient with a hereditary disorder that their child for example has a mutation in a specific gene, if the name is too comical, it's really not appropriate in that setting and some people can find it actually offensive or distressing. So, we have to take really maybe more of a wider context into account when we're naming humans genes. For fruit fly researchers, it's mainly about what's discussed in a scientific context and in the lab perhaps. But for us, we have to look at the wider picture because obviously, human genes impact human health.
Kat - When it comes to naming human genes, what sort of names do you choose and how do you reflect what the gene is like or what it does?
Elspeth - Okay, so we have a few criteria that we run through. Ideally, we would name a human gene based on a known function of the gene product. So for example, if the gene encodes for an enzyme with a known function then we would try and name it on the basis of that enzyme. If there's not a known function then we would start looking at homology. So, how related it is to other genes that are already known or maybe have a known function. As in the case of genes that are related to fruit fly genes, we would look at homologues in other species and see what was known about their functions. If there's nothing like that or not a high level of homology then we would start looking at the structure of the protein encoded by the gene. Does it encode any specific regions, usually called protein domains or motifs? And then we would perhaps bring that into the naming. And we'll also of course take into account any information that has been published or any information that a researcher has come to us saying, "We know something about this gene or this gene product." So really, a variety of criteria but our favourite criteria is to name based on a known function of the encoded gene product.
Kat - This is going to be a strange question but do you have any particular favourite gene names or gene names that really stick out for you?
Elspeth - You get fond of the ones that you remember naming! So, one way we like to name genes is grouping them into gene families. A gene family is usually related by sequence similarities. So, the genes are all related to each other and as a result, they quite often have related functions, not necessarily the same function, but something similar about their function. It's quite rewarding to take a group of genes that have been named very disperately and then contact lots of lots of researchers - because that is the key part of our job is actually contacting the researchers because obviously, they know a lot more about these genes and the encoded proteins than we do. So, we contact researchers and then we say, "This gene is a member of the family along with these other genes and we'd like to group them altogether and name them like that so that other researchers can pull out the whole family." And so, there's a few gene families that I've worked on in the past that it's quite rewarding then when you see them being published and people are using the new nomenclature. What's most important for us is that whatever name we decide upon will be used, because if the name is not used then it's kind of defeating the purpose. And the purpose is for everybody to be able to find a name in the literature in publications and then know exactly which gene somebody is talking about.
Kat - Do you think we will see a day when all the whimsical names will be replaced by three-letter acronyms?
Elspeth - Well certainly, for human - I don't know if all of them because some of them are not offensive or pejorative in any way. I don't know if all of them will go. I mean, Sonic Hedgehog - sometimes people say that Sonic Hedgehog has been replaced. It hasn't. I mean, Sonic Hedgehog is just so entrenched in the literature that I think we'd be kind of cutting off our nose to spite our face to get rid of it. But some of the other ones like Lunatic Fringe or something. Yeah, when it's actually potentially offensive, we really have to get rid of them. I think they'll probably stick around in Drosophila though because they don't see any reason to change them and in fact, they think our way of naming is very boring, too prosaic, but we like to think it's more informative. Less fun perhaps!
Kat - Elspeth Bruford from the HUGO Gene Nomenclature Committee.