There are a few reasons convergent evolution is really unlikely here. 1) This particular gene isn't the obvious solution - there are many, highly diverse antif

There are a few reasons convergent evolution is really unlikely here. 1) This particular gene isn't the obvious solution - there are many, highly diverse antifreeze proteins, not to mention other mechanisms of freeze resistance (glycerol production, for example). 2) Even if it were, the genetic code is redundant, meaning that there are often several 3-base codons that code for a given amino acid. So even if the exact amino acid sequence is what mattered, the odds of using the exact same coding to obtain that sequence is unlikely. 3) The similarity extends beyond the coding region. It includes stretches of DNA in between and flanking the gene's code itself. These are stretches of DNA that normally mutate at a much higher rate than the coding region itself, and they aren't under the selective pressure of making a working protein, so there's no real evolutionary explanation for how they'd end up so similar.