Today (well, I started yesterday but my nephew got in the way) I decided it was high time I made good on my personal promise to practice Jennifer Raff’s guide on how to read a scientific paper (link is on the menu bar at the top).
Why is this paper relevant? Well, following the probable recovery of DNA from a dinosaur bone, from 8 times longer ago than any hominid lived, then as far as I’m concerned pretty much anything can be considered possible! And that includes finding hominid DNA in fossil remains.
So, following Jennifer’s step-by-step advice:
1) Read the Introduction.
2) Identify the BIG QUESTION.
What is this whole scientific field trying to find out?
Can genetic studies of Old World monkeys (OWMs) tell us anything about human genetics (in either an evolutionary or biomedical context)?
3) Summarise the background in ~5 sentences.
Include: previous work done, its limitiations, any next steps.
- The dates divergence of OWM lineages (rhesus macaque/baboon) is similar (6-8 million years ago) to that of our own evolutionary divergence from chimpanzees.
- Lots of individual work has been carried out on the genomes of OWMs but it is isolated and uncollated.
- Non-human primate biomedical models have so far been used to investigate the only functional aspect of diseases/drugs.
- A full genomic model system needs comparative analysis of several taxa across all chromosomes including the sex chromosomes.
- All mechanisms of mutation should be considered, including methylation of DNA bases. (The other main way mutations arise is through imperfect replication of DNA every time a cell divides).
Now, when I started to move on, I noticed I had overlapped 3 into 4. Therefore:
- Extensive genetic studies of hominoids have already been performed, but reference to an outgroup is always prefereable in order to test general principles/hypotheses.
- The work done so far on OWMs has largely been independent and is uncollated although it has established evolutionary parallels between OWMs and hominoids.
- The biomedical aspect of using OWMs as non-human primate model systems has proven valuable but is currently limited to functional studies.
- The logical progression is to enable meaningful comparative genomic analysis for both the biomedical and evolutionary fields to complement the functional/anatomical data that already exists in a suitable format for comparative study.
4) Identify the SPECIFIC Question(s)
What are the aims of this particular paper?
Full genomic analysis of 4 taxa of OWM along with comparative study (to include humans) to establish the measure of genomic divergence, along with the rate of mutation which led to this. Does methylation cause mutations more often than other mechanisms of mutation?
But those aren’t very focused questions. Hell, the first question isn’t even worded as a question. So:
- How much genomic difference has accumulated between the subject taxa?
- What is the rate of genetic change that has occurred to achieve this?
- Does rate of change vary between autosomes and sex chromosomes?
- Does rate of mutation due to methylation differ from other mechanisms of mutation?
5) How will the authors address these questions?
By carrying out full genome scans and aligning sequences as best as possible for as much of the genome as possible, then compare them to look for differences which are due to mutations.
6) The Methods Section
So I didn’t draw pictures as suggested because I’m not much of a draw-er. Here’s the gist of it though:
The humongous strings of letters which represents the genomic DNA an organism was downloaded for a lot of individuals of OWMs and hominoids. 443 baboons, covering two species; 301 black-and-white colobus monkeys; 305 grivets.
All of these were lined up relative to the human genome to create chains of regions which match, separated by short stretches which don’t match.
I didn’t understand all this so I looked up some of the references and followed a trail of breadcrumbs to get to the bottom of it all (with a Famous Web Search Engine search in the middle).
But first, some background.
DNA is built from 2 strands wrapped around one another: Watson and Crick’s famous double helix. The strands stay together because the building blocks (nucleotides) pair with each other: A-T and C-G. Because DNA is used by all living things (plants, animals, bacteria and fungi), pairing happen can happen even if the strands are from different species providing the sequences are similar enough, which is the basis of this technique.
So, the computer programme analyses the sequence of each strand and assigns matching nucleotides a score of +1. For each stretch of DNA analysed, the score must be greater than +50, and it must cover at least 70% of the length of the corresponding human sequence. This creates the chain, and is carried out for each individual genome from each species that the researchers are investigating.
Fine, I drew pictures.
These were already available for human/rhesus, human/chimp human/orang-utan pairs.
The chains are then netted, which involves ordering the chains for each species comparison in order of decreasing % match to humans. Each chain is then analysed starting with the best-matching chain and any part of a chain that isn’t covered by the best match is then added to the best match chain to create the net.
After doing this you end up with a fairly huge stretch of very reliable and very well-matched DNA across all of the species. I don’t know why it’s called a net.
Methylation occurs on a C only where it is immediately followed by a G (a CpG site). The subsequent loss of nitrogen from a methylated C turns it into T, introducing a mutation. When this mutated strand is replicated during cell division, the second strand will now read CpA.
Subsequent copies of the CpA strand will read TpG.
This is a tell-tale signature of methylation-induced mutation and, by analysing all of the nets, it is possible to tell how many CpG sites there would have been in the common ancestor of each pair, and hence how regularly these mutations take place.
Specifically, the paper is looking for both types of mutation in non-repetitive, non-coding DNA because non-coding DNA is likely to be less susceptible to evolution by natural selection since it has no real function in the health of the animal.
7) Read the results
- Table 1: The X chromosome has diverged less than the autosomes in all species which were freshly analysed in this paper, i.e. data is not shown for chimp or orang-utan.
- Table 2: Comparing the differences in amount of divergence of the X chromosome and the autosomes shows that males are responsible for twice as many mutations than females (the male mutation bias). It is impossible to calculate a meaningful value for the male mutation bias for the OWMs because there is too much variation (See the range of values in Table 2).
- Figure 1: We can, however, see that the rate of evolution is different between the different species of OWM. This finding is, scientifically, ‘highly significant’.
(p-value<10^-5. The probability that this result is down to chance is less than 0.0001. On a probability scale of 0 – 1 that is vanishingly small. We can be incredibly confident that this result is correct).
- Table 3: The baboon/rhesus pair has undergone 30% more substitutions than the human/chimp pair when looking at non-CpG sites, i.e. when methylation is not responsible.
When you look at CpG sites (where methylation is responsible) the number of mutations is the same (remember that the baboon/rhesus lineages split at the same time as the human/chimp lineages).
8) Do the results answer the specific questions? What is the significance of this information?
How much genetic difference has accumulated between taxa?
Read: how much have the different taxa diverged? Table 1 tells us this, as well as the order in which the divergences took place. (Lower number means more recent divergence).
What has been the rate of genetic change?
Figure 1 tells us the average number of mutations per nucleotide that have taken place since each lineage split. For the human/OWM split this is about 0.04 mutations per nucleotide, or one mutation every 400 nucleotides. This is the average difference of the human genome when compared to each of the OWM genomes.
This tells us how many mutations have taken place and allows us to work out the date of the divergence, assuming mutations occur at a constant rate.
Does mutation rate vary between autosomes and the X chromosomes?
Yes (see Table 1). This is because the autosomes exchange corresponding stretches of DNA within their pair each time a cell divides, changing the exact sequence each chromosome possesses. This does not happen as often for the X chromosome because it cannot exchange material with the Y chromosome in males, leading to a lower rate of change.
Does mutation rate vary between CpG and non-CpG sites?
Yes (see Table 3). This may mean that if we use both types of mutation to calculate dates of lineage divergence we may not get the true date.
9) Read the discussion.
What is the authors’ interpretation? Do I agree? Are there any alternatives? Are weaknesses in the study identified? Have I spotted anything the authors didn’t?
The authors’ interpretations:
- OWM genomic data is readily avaible and of high enough quality to be used as a useful model system. It is possible to distinguish common patterns between OWMs and hominoids, such as the male mutation bias and methylation.
- The male mutation bias is found to be generally lower in primates than previously thought, although it does vary between primate species. This is thought to be down to generation lenght. The earlier in life the first reproduction takes place, the fewer times the sperm-producing cells have divided and have accumulated fewer mutations.
- Baboons evolve more slowly than any of the other OWMs investigated. They believe that this is likely down to specific events in evolutionary history of the OWMs, such as where ancestral individuals may have interbred with the ancestors of another OWM lineage before they fully speciated.
I wonder if this may have something more to do with generation time and the number of cell divisions and hence chromosome recombinations, in a similar manner to male mutation bias? The authors do not explicitly mention this or rule this out.
Generally then, I agree. Although I do wonder why they haven’t included any data on chimps and orang-utans in the paper. It doesn’t say either in the paper itself or in the Supplementary Online Information (SOI).
They go on to say:
“[…] the fact that substitution rates vary substantially and
significantly between even closely related lineages
make us ask whether we can confidently infer divergence
times and other evolutionary events from
This is a very good point, identifying weakness in their own study, although they do note afterwards that rates of mutation at CpG sites are much more similar and therefore likely to be more useful.
Again, I agree with this interpretation.
10) Read the Abstract
Does it match the interpretations in the discussion? Does it match MY interpretations?
The numbered points are quoted directly from the abstract, bullet points are my thoughts.
“(i) Genomic divergences among several Old World monkeys mirror those between well-studied hominoids.”
- This isn’t explicitly shown by this particular study. I hope they cross referenced their data to that of other studies. The fact that they did this isn’t stated anywhere in the article itself or the SOI.
“(ii) The X-chromosome evolves more slowly than autosomes, in accord with ‘male-driven evolution’. However, the degree of male mutation bias is lower in Old World monkeys than in hominoids.”
- I agree with their interpretation of X chromosome evolution as being due to male mutation bias.
- Their data do also suggest that OWMs experience less male mutation bias than hominoids, and this is likely due to shorter generation times.
“(iii) Evolutionary rates vary significantly between lineages. The baboon branch shows a particularly slow molecular evolution. Thus, lineage-specific evolutionary rate variation is a common theme of primate genome evolution.”
- This is quite evident in their Results section.
“(iv) In contrast to the overall pattern, mutations originating from DNA methylation exhibit little variation between lineages.”
- I accept this.
The abstract concludes:
“Our study illustrates the
potential of primates as a model system to investigate genome evolution, in particular to elucidate molecular mechanisms of substitution rate
I think this is a fair assessment as the paper demonstrates that OWM genomic data is abundant and of high quality, and that robust parallels can be drawn to hominoid genomic data.
I know that was long-winded, but many congratulations if you got to here. Brownie points are heading your way as you read this.
If you read anything that you don’t agree with please mention it in the comments and I’ll be more than happy to correct myself. Or you.
Wilson-Sayers M., 28th June, 2012. Accessible research: Male mutation bias. mathbionerd. Available from: http://mathbionerd.blogspot.co.uk/2012/06/accessible-research-male-mutation-bias.html. [Accessed 17th September, 2013].