How similar are phylogenetic trees?

[Jerry Johnson]

Hi all,

I read that phylogenetic trees are one of the strongest evidence of evolution, because we get a very similar trees for different genes and proteins. But I can’t find anywhere ANY concrete/specific numbers. How many phylogenetic trees where constructed so far? 1000? 50,000? a million? How similar are they to one another? 99%? 95%? 80%?

I would really like to see some numbers, where can I find them?

I’m asking this because I came across a video claiming that phylogenetic trees do NOT support evolution:

https://www.youtube.com/watch?v=lk1gDk1wGhQ&t=5m52s

So, if you have some numbers it will really help.

Thanks.

 

[CharonY]

I think it will be hard to find total numbers. Phylogenetic trees are a standard technique used for tons of different questions . But I think that is beside the point. First the assumption that all trees must be the same is not true. If it was the case reconstructing them would be a waste of time in many cases. Rather, what you expect to see is that for example in very conserved genes (i.e. genes that highly selected for, because they are fundamental) the trees will more closely follow species relationship (as we assume even pressures). On the other hand, genes that are not under strong pressures we expect to see higher divergence. I.e. because there is no strong selection to maintain them, they are more prone to mutations. 

[Jerry Johnson]

22 hours ago, CharonY said:

I think it will be hard to find total numbers. Phylogenetic trees are a standard technique used for tons of different questions . But I think that is beside the point. First the assumption that all trees must be the same is not true. If it was the case reconstructing them would be a waste of time in many cases. Rather, what you expect to see is that for example in very conserved genes (i.e. genes that highly selected for, because they are fundamental) the trees will more closely follow species relationship (as we assume even pressures). On the other hand, genes that are not under strong pressures we expect to see higher divergence. I.e. because there is no strong selection to maintain them, they are more prone to mutations. 

Thanks, but still, I expect that if you'll take several represent animals (monkeys, humans, dogs, cats, zebras, horses, birds, bats, squirrels, fish, whales, dolphins ...) and you'll build several phylogenetic trees for them based on several different genes and proteins, then you should get very similar trees. If not, then you can't say that phylogenetic trees are evidence of evolution.

So my question is still relevant, how close are the different trees to one another?

 

[CharonY]

1 hour ago, Jerry Johnson said:

nd you'll build several phylogenetic trees for them based on several different genes and proteins, then you should get very similar trees.

No, if evolution is true that is exactly what you do not expect. One aspect of evolution is that selective pressures shape the evolution of genes. I.e. you can have maintaining presssures, forcing genes not to changes. This is true for example true for essential genes, say for respiratory genes. OTOH, areas that are under weak or no selective pressure are expected to diverge more strongly. They are likely to be reasonably close among close lineages but the farther you go the more uncertainty you get. IOW if you build phylogenetic trees using conserved sequences, you get a tree that will follow lineages very closely. If you use areas that are highly variable, you will other shaping factors (such as drift). Again, this is what we expect if evolution happens. Your scenario (i.e. that phylogenetic trees will also resolve the same way regardless of loci being analyzed) will only happen if there is in fact no evolution, i.e. the genomes are either static or all change at exactly the same rate.

[Jerry Johnson]

13 hours ago, CharonY said:

No, if evolution is true that is exactly what you do not expect. One aspect of evolution is that selective pressures shape the evolution of genes. I.e. you can have maintaining presssures, forcing genes not to changes. This is true for example true for essential genes, say for respiratory genes. OTOH, areas that are under weak or no selective pressure are expected to diverge more strongly. They are likely to be reasonably close among close lineages but the farther you go the more uncertainty you get. IOW if you build phylogenetic trees using conserved sequences, you get a tree that will follow lineages very closely. If you use areas that are highly variable, you will other shaping factors (such as drift). Again, this is what we expect if evolution happens. Your scenario (i.e. that phylogenetic trees will also resolve the same way regardless of loci being analyzed) will only happen if there is in fact no evolution, i.e. the genomes are either static or all change at exactly the same rate.

So what's the point? If each gene (or protein) gives you a different phylogenetic tree, then how exactly does it shows us the relationship between the species? How does it prove evolution?
 

 

[delboy]

7 hours ago, Jerry Johnson said:

So what's the point? If each gene (or protein) gives you a different phylogenetic tree, then how exactly does it shows us the relationship between the species? How does it prove evolution?
 

 

 

An evolutionary biologist would no doubt know how to extract the relevant information to construct a tree. As CharonY said, certain genes indicate the species relationship more reliably, so use those ones.

Phylogenetic trees are not constructed to prove evolution. Scientists simply accept its truth (unless and until it may be happened to be proven wrong). If you wish to use them as evidence you simply have to quote CharonY's extra detail - which actually contains information, unlike the video, which simply states (false) things without providing any evidence.