Of Terms in Biology: Ortholog

Source.

The terms ortholog and paralog have wide acceptance. Yet, we find that they are used in nuanced ways by different people. Sure enough, here is some of what the Internet comes up with:

• ortholog: a gene from a disparate species, so similar in genetic sequence that it is assumed to have originated from a single gene in a common ancestral species.


• ortholog: either of two or more homologous gene sequences found in different species.


• orthologous: exhibiting orthology; having been separated by a speciation event.


• orthologs: homologous genes that result from speciation events (as opposed to paralogs, which result from duplication and subsequent divergence in a lineage).


• orthologous: groups of genes or proteins from different organisms that have the same function are said to be orthologous.


• orthologous: genes in different species are orthologous when their phylogeny is identical to the phylogeny of the species.

We deleted the ones that seemed like duplications (paralogs?). We find the first one most to our liking. Which do you think is most useful?

Comments

[Rhea Miller]

lol. Way to point out one of the hardest parts of graduate school. While learning NEW information....where do you get it?!?!? when do you trust it?!?!? how do you know that you are getting the best answer?!?!?!

...what a conundrum.

[Matt Hudson]

I agree with Jonathan Badger that to most bioinformatics types "orthologous" is synonymous with "shared function" as well as "shared lineage". However I think the two are both implied by the evolutionary definition, since "function" here is in the broad sense of biochemistry and structure, not the narrow sense of fitness in a particular organism. Thus, if duplication and subfunctionalization has occurred, the gene is no longer a strict ortholog in either sense.

It is certainly implicit to most bioinformatics people that orthology implies common ancestry. Homeoplasy would never be confused with orthology.

[Jonathan Badger]

Well, "groups of genes or proteins from different organisms that have the same function are said to be orthologous" is the way most bioinformaticists use the term, even though most evolutionary biologists are horrified by it.

I'm not sure it is an issue of "usefulness"; both the evolutionary and bioinformatics meanings are useful -- but as the evolutionary biologists had the term first, I suppose their meaning of traits/genes separated by a speciation event has priority, and the bioinformaticans should invent a new term.