Evolution problem

[Podkayne]

I have a friend who, despite seeming normal in most other respects, is a creationist. We've long ago been through all the PRATTs before, but today he caught me with a real puzzler, that I simply can't hand wave away;

Choanoflagellates are thought to be our closest living single celled relative, and its not unreasonable to think they are probably pretty similar to our most recent single celled common ancestor. The genome of  Monosiga brevicollis has been sequenced and has 41.6 million base pairs. It has about 9200 genes in its genome. The human genome encodes roughly 23 000 genes, in approximately 2.9 billion base pairs. Without even considering polygenic traits, and hypernucleic genes and DNA, and imagining there is a direct progression between choanoflagellates and humans, without considering the ascendance and disappearance of various traits along the way, the question is just how many distinct speciations occur between the MRsingle celled CA and us, how many instances of syntax appropriate, information adding mutations would be involved in that progression, and, lastly, what is the minimum required rate of stable mutation per generation required to go from choanoflagellates to us?

Now I'm not a molecular biologist by any stretch of the imagination, much less anything like a mathematician, but I have to admit, I keep trying to come up with some educated guesses based on what I do know, and I keep roughly working out that for Neodarwinian evolution to account for humans from choanoflagellates, we're missing a couple tens of billions of years in which it could have occured.

So, any one want to think about this one and throw some figures at the wall?

[m52nickerson]

Okay let's start with some base numbers.

Eukaryotes evolved about 1.85 Ga (billions of years ago)

Modern Humans came about 200,000 years ago.

So we have 1.848 Ga to work with.

[Podkayne]

Okay let's start with some base numbers.

Eukaryotes evolved about 1.85 Ga (billions of years ago)

Modern Humans came about 200,000 years ago.

So we have 1.848 Ga to work with.

Right. Doesn't seem like enough time, but I'm glad someone else will have a think about it now!

[m52nickerson]

Right. Doesn't seem like enough time, but I'm glad someone else will have a think about it now!

How does 1.8 billion years not seem like enough time?

Early on mutations would be happening quite fast.  Each generation is going to come about in a few hours to a day at the most.  Even when we get to higher forms, for much of the timeline we will be getting new generations once a year at least.  It will not be until the very end in which each generation comes about over multiple years ad the rates of mutation slow way down.

Even with that slow down it is thought that Human and Chimps diverged only 7 million years ago. 

[Shane for Wax]

I think these links might aid you:

The genome of the choanoflagellate Monosiga brevicollis and the origin of metazoans
Jumping off point for following links to various evolutionary articles

And though I don't like Dawkins, his Ancestor's Tale does a good job of tracing things evolutionary biology wise. I'd recommend borrowing from your local library though. It's not a cheap book. And since I can't imagine it being something you need constantly...

Also remember, Evolution may work 'slowly' but sometimes it isn't as slow as people might think.

[Lithp]

1. You can't compute the probability of "something evolving the way it did." So if a Creationist says that, they're bullshitting, quite frankly.
2. Sometimes, it's difficult to comprehend how ridiculously huge these numbers are. A billion is a a thousand millions. There are a thousand thousands in a million. Human civilization spans roughly 6-10 thousand years. A generation, for humans, is approximately 30 years. And, as Nickerson noted, certain links in the chain had much shorter generational periods.
3. Some of these numbers might be a bit off. We can be certain about the age of the earth. The age at which eukaryotes evolved? Might have occurred earlier than the current estimation.
4. Chromosome duplication. It causes problems in humans, but not in some other organisms. Plants do it all of the time. So forget the genome multiplying over millions of years. In the right circumstances, it can do it in a few generations.
5. There are some issues in your friend's question. I'm not supposed to take into account natural selection of favored traits? I'm supposed to apply a constant rate of change that applies to all organisms? Even if I can do that, he shouldn't be surprised if it doesn't reflect reality very well, because that's not how evolution works.
6. We might just plain be missing some of the early information. The organisms you're talking about aren't good with the whole fossilization thing, & genetics is relatively new.

Hope that helps.

[Podkayne]

Right. Doesn't seem like enough time, but I'm glad someone else will have a think about it now!

How does 1.8 billion years not seem like enough time?

Early on mutations would be happening quite fast.  Each generation is going to come about in a few hours to a day at the most.  Even when we get to higher forms, for much of the timeline we will be getting new generations once a year at least.  It will not be until the very end in which each generation comes about over multiple years ad the rates of mutation slow way down.

Even with that slow down it is thought that Human and Chimps diverged only 7 million years ago.

Well how many stable mutations and speciation events do you come up with as  being necessary? On additional base pairs alone, we have 2858400000  additional ones, let alone ones that change from present to absent, or vice versa, between the collared cell and us. So for that alone, we'd require at least two additional base pairs each year, on average, in stable mutations. Unless I carried a zero wrong... anyway, doesn't that seem rather a lot, especially since there's a lot more going on here than just adding on a base pair at a time.

I agree 1.5 billion years is a really long time, its just, on the face of it, it doesn't seem like enough time for natural selection of traits arising from random mutations to achieve what they have. Maybe I'm way off base in my thinking about what sustainable average rates of speciation are. is there a molecular geneticist in the house?

[m52nickerson]

We are going from 2.9 million base pairs to 41.6 million base pairs.  So 38.7 million base pairs.  I don't know where you got your 2.8 billion base pair number from.

It took 1.848 Ga to do it so....

38,700,000/1,848,000,000 = 0.02094

So we average a change of 0.021 base pairs each year, or a 1 new base pair every five years or so.

[Podkayne]

We are going from 2.9 million base pairs to 41.6 million base pairs.  So 38.7 million base pairs.  I don't know where you got your 2.8 billion base pair number from.

It took 1.848 Ga to do it so....

38,700,000/1,848,000,000 = 0.02094

So we average a change of 0.021 base pairs each year, or a 1 new base pair every five years or so.

2.9 billion... base pairs in the human genome, if I recall correctly

The human genome is made up of DNA, which has four different chemical building blocks. These are called bases and abbreviated A, T, C, and G. In the human genome, about 3 billion bases are arranged along the chromosomes in a particular order for each unique individual. To get an idea of the size of the human genome present in each of our cells, consider the following analogy: If the DNA sequence of the human genome were compiled in books, the equivalent of 200 volumes the size of a Manhattan telephone book (at 1000 pages each) would be needed to hold it all.

It would take about 9.5 years to read out loud (without stopping) the 3 billion bases in a person's genome sequence. This is calculated on a reading rate of 10 bases per second, equaling 600 bases/minute, 36,000 bases/hour, 864,000 bases/day, 315,360,000 bases/year.

Storing all this information is a great challenge to computer experts known as bioinformatics specialists. One million bases (called a megabase and abbreviated Mb) of DNA sequence data is roughly equivalent to 1 megabyte of computer data storage space. Since the human genome is 3 billion base pairs long, 3 gigabytes of computer data storage space are needed to store the entire genome. This includes nucleotide sequence data only and does not include data annotations and other information that can be associated with sequence data.

As time goes on, more annotations will be entered as a result of laboratory findings, literature searches, data analyses, personal communications, automated data-analysis programs, and auto annotators. These annotations associated with the sequence data will likely dwarf the amount of storage space actually taken up by the initial 3 billion nucleotide sequence. Of course, that's not much of a surprise because the sequence is merely one starting point for much deeper biological understanding!

Contributions to this answer were made by Morey Parang and Richard Mural formerly of Oak Ridge National Laboratory; and Mark Adams formerly of The Institute of Genome Research.
http://www.ornl.gov/sci/techresources/Human_Genome/faq/faqs1.shtml

From the human genome project FAQ.

And 41.6 million is the number to base pairs in the hypothetical precursor organism, you seemed to read my post as saying thats the number in humans.

I could have miss placed a zero there, sorry.

[m52nickerson]

Okay, so I was all screwed up, lets do this again.

Going from 41.6 million base pairs to 2.9 billion base pairs, so their is your 2,858,400,000.

So you get an average of 1.54 base pairs per year.

Yes that seems like a lot, but that is just an average. Again rates of change would have been much higher at the beginning.

The rates of mutation for eukaryotes is about 0.003 mutations per genome per generation.  You can easily have each new generation within hours.  Rates are even higher in viruses which can be 10 to the 6th.  Plus we have no idea what mutation rates were like so many years ago.

It may not seem possible, but that does not change the other evidence we have.

[Podkayne]

Okay, so I was all screwed up, lets do this again.

Going from 41.6 million base pairs to 2.9 billion base pairs, so their is your 2,858,400,000.

So you get an average of 1.54 base pairs per year.

Yes that seems like a lot, but that is just an average. Again rates of change would have been much higher at the beginning.

The rates of mutation for eukaryotes is about 0.003 mutations per genome per generation.  You can easily have each new generation within hours.  Rates are even higher in viruses which can be 10 to the 6th.  Plus we have no idea what mutation rates were like so many years ago.

It may not seem possible, but that does not change the other evidence we have.

Don't get me wrong, I believe in common decent with modification, but theres a lot more to the changes betweem them and us than merely adding an extra codon every year and a half. I was trying to simplify the problem.

I'm not suggesting evolution is wrong, just that on the face of it, in the allowed time, it looks like there may be some other unidentified principle taking effect. This is all sort of academic, since we'll never know the actual properties of the genomes in question.

[StallChaser]

Mutation rates tend to be higher in simple organisms, because the individual cost of a bad mutation (most of them) isn't as high.  It becomes worth having a higher mutation rates for the potential of beneficial mutations, that isn't so much the case for larger organisms.  That would help explain the seeming discrepancy, if you were using the mutation rate of a human for a single-celled organism.  Also, a single mutation can potentially insert a large chunk of DNA, so it's not like you can infer a number of base pairs added by the mutation rate alone.

But as for a possible real answer, the tree of life web project has mapped out a lot of the evolutionary paths, so you might be able to trace a path from Choanoflagellates to H. Sapiens.

[Podkayne]

Mutation rates tend to be higher in simple organisms, because the individual cost of a bad mutation (most of them) isn't as high.  It becomes worth having a higher mutation rates for the potential of beneficial mutations, that isn't so much the case for larger organisms.  That would help explain the seeming discrepancy, if you were using the mutation rate of a human for a single-celled organism.  Also, a single mutation can potentially insert a large chunk of DNA, so it's not like you can infer a number of base pairs added by the mutation rate alone.

But as for a possible real answer, the tree of life web project has mapped out a lot of the evolutionary paths, so you might be able to trace a path from Choanoflagellates to H. Sapiens.

Cool link, thanks

[Lithp]

I'm not suggesting evolution is wrong, just that on the face of it, in the allowed time, it looks like there may be some other unidentified principle taking effect.

Yes, as I said, there is a lot more to mutation than "insert a new codon here."

[Quasirodent]

ALSO, keep in mind how many individuals are involved in each generation.  The line from one species to the next isn't made up of a direct lineage from one single animal to the next, but an average over millions or billions of individuals over entire populations.

If I'm not mistaken, not only would generations from birth to reproductions be much shorter in simpler organisms, but they would have been staggered and overlapping, as the parent would continue to reproduce while its earlier offspring matured.