Re: Genetic Similarity

On Apr 23, 4:22 am, "alwaysaskingquestions"
<alwaysaskingquesti...@xxxxxxxxx> wrote:
There's been a fair bit of discussion here about the DNA similarity between
different species, particularly between chimps and humans.

I'm not knowledgeable about genetics so I hope this is not a stupid
question, but can someone explain to me in layman's terms, what percentage
of DNA (if it can be measured in that way) can be regarded as common to
*all* species?

Not really anything that you would write home about. As you look at
more and more distantly related species it gets harder and harder to
determine what DNA they still share for the simple reason that more
change has happened. We get around this problem by looking at
specific genes that have been conserved enough that we can tell that
they are still the same gene. So we get answers for those specific
genes and it is no surprise that genes vary in how similar they are in
the various species. Certain genes are more conserved than others.
It turns out that the various degrees of conservation present
limitations to how we can evaluate the differences.

There are really conserved genes that are basically useless for
phylogenetic comparison because they are so conserved. Take histone
genes. There is only a single amino acid change between plants and
multicellular animals in one of them. This means that we can't use
the sequence to tell anything about ancient metazoan lineages because
the only places the DNA has changed are at degenerate sites that do
not change the amino acid sequence and we can infer that such sites
must have been hit by mutation multiple times during the evolution of
metazoans so they are worthless in evaluating phylogenies. So even
though the DNA sequence is similar it holds no really valuable
phylogenetic information.

There are other genes that vary a lot in their amino acid sequence,
but if they vary too much they also aren't good information sources.
It turns out that you have to pick genes that are evolving at certain
rates to get the most accurate estimates of the associations between
the various species.

Something like hemoglobin varies at a rate that makes using it out
past around 300 million years problematic and error prone. Though
cytochrome c is shorter it can be used to go further back with more
accuracy because it is more conserved than globin. You'd still like
to get more sequence to get the most accurate picture.

You just have to look at it this way, once 10% of the DNA sequence has
changed, given a random distribution the next mutation has a 10%
chance of hitting a site that has already changed. This would lead to
errors in estimation of how related the sequences are. You couple
this fact with the degenerate codon sites and the fact that they
"saturate" (are hit multiple times) more rapidly than nondegenerate
sites and you begin to get an idea of what goes into these types of
analyses. For more distant relationships we can remove degenerate
codon sites from the analysis so that source of error can be
minimized.

For species as closely related as chimps and humans we can still use
degenerate codon sites because they obviously are far from
saturation. Gene coding regions only vary by less than 1% between
chimps and humans. When you get out to humans and mice where the same
gene may vary by 20% in its DNA sequence you can begin to worry, and
if you are dealing with some fish and humans you would be dropping the
degenerate sites from the analysis.


Also, do we have any idea of the scale of change in DNA in Man today from
his ancient ancestors? I realise that we don't have ancient DNA to work
with, but I'm wondering if their has been any extrapolation done.

We can make this estimate, but prior to just a couple of weeks ago we
didn't have enough data to nail any such estimate. With the recent
publication of a monkey genomic sequence to compare with the human and
chimp genomic sequence we can now make fairly accurate estimates on
rates of change between the two lineages on a genome wide scale
instead of a few genes.

It works this way, just having two sequences doesn't tell you what
changes occurred in each lineage it only tells you how different the
two sequences are, but when you get a third sequence of an outgroup
species to compare the two more closely related species to you can now
get an estimate of which changes occurred in which lineage. Say
humans have a sequence AGTCTTTCCGGAA and chimps have a sequence
AGCCTTTCTGGAG. The sequences vary at three positions, but you have no
information whether the mutations occurred in chimps or humans. You
now get a monkey sequence that is AATCTCCTGAAA. You see that monkey
sequence varies from both chimp and human by more substitutions, but
you also get some information about the possible ancestral sequence.
You find that two of the three differences between chimps and humans
likely occurred in the chimp lineage because monkeys share the human
sequence at those positions.

Naturally you would like more than just a three way comparison, but
three is the minimum needed for such an analysis.

It turns out that certain chimp genes may be changing faster than the
corresponding human genes. They can also infer if the changes were
due to random drift or selection and they claim selection is
responsible. Science isn't trying to snow people, we have been
dealing with this data for decades and it all tells a consistent
story. Challenging the concept of common descent is just as bad as
claiming that the earth is flat.

Ron Okimoto

.

Relevant Pages

  • Re: Chimpanzee-Human DNA comparisons
    ... researchers reported "only" a 95% similarity between chimps and humans, ... each unmatched sequence. ... Deletion and insertions are usually not ...
    (talk.origins)
  • Re: Chimpanzee-Human DNA comparisons
    ... researchers reported "only" a 95% similarity between chimps and humans, ... sequence stretches are often involved in deletions and insertions. ... Deletion and insertions are usually not ...
    (talk.origins)
  • Re: Some more Bull****! Whats the 1.2-5% dna Diff! (Chimp to man)
    ... lions are closely related to humans. ... Percentage of sequence similarity, on the other hand, is highly ... "If all carbon based life has DNA; Human DNA is closest to chimp DNA ... that one simple datum -- degree of sequence similarity between chimps ...
    (talk.origins)
  • Re: Genetic Similarity
    ... genes and it is no surprise that genes vary in how similar they are in ... the sequence to tell anything about ancient metazoan lineages because ... For species as closely related as chimps and humans we can still use ...
    (talk.origins)
  • Re: What makes us human?
    ... Chimps and humans each have about three billion nucleotides in their ... The discovery of genes ... preserves favourable mutations. ...
    (talk.origins)
Loading