Match the gene's speed to the age of the split
- The idea
- Predict
- See it
- Why
- Check
The idea
“More differences make a clearer tree, so the fastest-changing gene is always best.”
More differences sound like more evidence. Over a long span, a fast sequence keeps changing the same sites, and its count stalls.
The short version
The mistake: More differences make a clearer tree, so the fastest-changing gene is always best.
What's actually true: A marker must change enough to separate the groups, but not so fast that it stops tracking time.
- A molecular clock dates a split by counting DNA differences between lineages. It only works while that count keeps rising.
- Noncoding DNA under little selection changes fast and keeps hitting the same sites. A later change can overwrite or undo an earlier one. This is saturation.
- Unrelated DNA sequences still match at about 1 site in 4 by chance, so a saturated sequence stalls near 75% different.
- A slow gene, like a ribosomal RNA gene, has the opposite problem. It barely changes over a recent split.
On the exam: Use a fast marker for a recent split and a slow one for an ancient split.