tag:blogger.com,1999:blog-37148773.post3956712453652025149..comments2024-03-19T00:24:23.577-04:00Comments on <center>Sandwalk</center>: Revisiting the genetic load argument with Dan GraurLarry Moranhttp://www.blogger.com/profile/05756598746605455848noreply@blogger.comBlogger84125tag:blogger.com,1999:blog-37148773.post-92073512722613834542023-06-05T17:55:57.712-04:002023-06-05T17:55:57.712-04:00Michaeljf said:
"Joao (and Larry) - I am work...Michaeljf said:<br />"Joao (and Larry) - I am working my way through this Galeota-Sprung et al and comparing to Graur (2017)."<br /><br />Thank you for your comment, Michael! I revived this comments as I read "What's in your genome". I think I brought Larry's attention to Galeota-Sprung paper, and I'm happy this is somewhat in the book. Larry wrote:<br /><br />"Using these values, Dan Graur estimated that at least 75 percent of our<br />genome has to be junk, and it’s likely that the actual amount of junk DNA is<br />closer to 90 percent. However, a more recent analysis shows that calculating<br />the fraction of junk DNA is a lot more difficult than Graur thought and<br />certainly a lot more complicated than the simplistic calculations that I<br />presented earlier[4]."<br /><br />Footnote 14 refers to Graur (2017) and Galeota-Sprung et al. (2020). <br /><br />:)Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-79191635369324297112022-02-01T17:54:47.959-05:002022-02-01T17:54:47.959-05:00Joao (and Larry) - I am working my way through thi...Joao (and Larry) - I am working my way through this Galeota-Sprung et al and comparing to Graur (2017). I agree they don't make a substantive comment on the % functional ("f") in either their abstract, intro or conclusion. However, they do say some interesting things in the middle. They reconcile their model with Graur's (note: they believe Graur didn't correctly include a factor of 1/2 to account for diploid vs haploid counts - but they show the corrected interpretation of Graur to have a better apples-apples comparison). Then they do make this statement: "Our interpretation of the quantity w_max/w_bar is more liberal than Graur’s: We do not interpret it as mean requisite fertility, because we are not using a pure viability selection model, but as the fitness of the fittest individual. Thus, our interpretation of the approach of Graur yields somewhat higher possible values for f than occur in Graur (2017), as shown in the bottom row of table 1, but still almost certainly no higher than f = 0.10 given the parameter values in (3)". And the "parameter values in (3)" are their estimates earlier in the paper of the main inputs, namely, size of genome, probability of a deleterious mutation at a site, impact of a mutation on fitness ("s") and steady size of human population (conservative at 1 billion). So I find this interesting, because they have no real issue with the input values that reconcile to Graur (save the 1/2 correction) and they come up with an upper limit of f = 0.10. With the correction applied to Graur's paper this would be Graur's own table figure of 0.20 (and note Graur had 0.15 as the upper limit). So seems coming in around the same value, ie 0.80 or more = non-functional. With any apologies as I'm a layperson with respect to genetics/biology but was following the rabbit trail of people trying to rebut Casey Luskin (an effort which I support, but don't have the full training on!)michaeljfhttps://www.blogger.com/profile/02089997822691949943noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-47711622946590839442020-05-08T01:25:49.283-04:002020-05-08T01:25:49.283-04:00Also, I just became aware that Ben Garleota-Sprung...Also, I just became aware that Ben Garleota-Sprung briefly explained their argument here: https://twitter.com/SprungBen/status/1249800322236723200Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-37093248320024268832020-05-08T01:18:56.632-04:002020-05-08T01:18:56.632-04:00Larry, did you see that somebody on EN wrote somet...Larry, did you see that somebody on EN wrote something about Garleota-Sprung et al. paper? The claim is the following: "Paper Shows that “Mutational Load” Arguments Don’t Refute ENCODE". <br /><br />https://evolutionnews.org/2020/05/paper-shows-that-mutational-load-arguments-dont-refute-encode/Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-9801623227375942722020-04-14T11:49:06.164-04:002020-04-14T11:49:06.164-04:00Thanks. I'm still not clear about what the pro...Thanks. I'm still not clear about what the problem is with mutational load arguments.Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-55726306476759968122020-04-13T16:35:28.625-04:002020-04-13T16:35:28.625-04:00Larry, here is what Ben Garleota-Sprung commented ...Larry, here is what Ben Garleota-Sprung commented on their paper:<br /><br />https://twitter.com/SprungBen/status/1249790349725368321<br /><br /> Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-79022585627326288402020-04-13T12:39:40.003-04:002020-04-13T12:39:40.003-04:00I asked Dan on Twitter if he has some comments on ...I asked Dan on Twitter if he has some comments on Galeota-Sprung et al. This is what he said:<br /><br />"I was one of the two reviewers of this paper and I recommended publication without revision. I don't tend to fall in love with my own papers." <br /><br />You say "I find it very annoying that they avoid making any substantive comment on the amount of functional DNA in a genome". And I totally agree. They say mutational load is not that limiting, however they decline from discussing further on the probable amount of junk. Well, at least they seem not to deny that possibly a large fraction of our genome is junk. <br /><br /><br /><br />Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-65159903879402106582020-04-13T12:28:23.122-04:002020-04-13T12:28:23.122-04:00I don't know if Graur has commented on that pa...I don't know if Graur has commented on that paper. I just read the Galeota-Sprung et al. paper and I find it very annoying that they avoid making any substantive comment on the amount of functional DNA in a genome. Surely as experts they have some idea about reasonable selection coefficients? Also, it's not clear to me why the fitness of the most fit individual is so important. Perhaps some other expert can take a look at the paper? Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-8361727039569089392020-04-13T00:24:24.867-04:002020-04-13T00:24:24.867-04:00Larry, do you know if Graur responded or commented...Larry, do you know if Graur responded or commented on the following paper by Galeota-Sprung et al., 2019?<br /><br />Mutational Load and the Functional Fraction of the Human Genome<br /><br />"We find that the functional fraction is not very likely to be limited substantially by mutational load, and that any such limit, if it exists, depends strongly on the selection coefficients of new deleterious mutations."<br /><br />https://academic.oup.com/gbe/article/12/4/273/5762616<br />Joãohttps://www.blogger.com/profile/06897303523201917516noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-33375670500709327662019-11-18T19:20:29.861-05:002019-11-18T19:20:29.861-05:00This seems relevant:
"We stress that we, in ...This seems relevant:<br /><br />"We stress that we, in this work, take no position on the actual proportion of the human genome that is likely to be functional. It may indeed be quite low, as the contemporary evidence from species divergence and intraspecies polymorphism data suggests. Many of the criticisms of the ENCODE claim of 80% functionality (e.g. Doolittle 2013, Graur 2013) strike us as well-founded. We wish only to point out that an argument from mutational load does not appear to be particularly limiting on f."John Harshmanhttps://www.blogger.com/profile/04478895397136729867noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-67935277944109427412019-11-18T15:03:14.049-05:002019-11-18T15:03:14.049-05:00I look forward to reading the paper after it has b...I look forward to reading the paper after it has been reviewed and published. Please let me know when that happens.Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-20579514886228484032019-11-16T17:33:07.292-05:002019-11-16T17:33:07.292-05:00You and Graur might be interested in this recent p...You and Graur might be interested in this recent preprint (https://www.biorxiv.org/content/biorxiv/early/2019/09/30/785865.1.full.pdf)Elihttps://www.blogger.com/profile/12570010209963549390noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-76177244488735633282017-08-28T14:58:34.042-04:002017-08-28T14:58:34.042-04:00David,
You seem to be claiming that a DNA sequenc...David,<br /><br />You seem to be claiming that a DNA sequence must be essential in order to affect fitness. What sort of "local, mechanistic/metabolic role" would not affect fitness?<br /><br />While we're here, what's wrong with "junk"?John Harshmanhttps://www.blogger.com/profile/04478895397136729867noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-62673560258971018192017-08-28T13:19:26.993-04:002017-08-28T13:19:26.993-04:00>> However it is not a definition that most ...>> However it is not a definition that most biologists will intuitively embrace<br /><br />>What is wrong with a definition of function that relates it<br />>to the genetic fitness of an organism or its offspring?<br />>This seems to me to be the only obvious definition.<br /><br />I disagree that there's anything "obvious" about function equating to fitness. I would say that among most scientists, the more common conception of "function" does not entail essentially, especially to the whole. A protein product of a gene may very well play a local, mechanistic/metabolic role without it having any bearing upon the reproductive fitness of the organism. In this case, it would be common for many to say that this gene still had a "function" without then going on to make any presumptions about that gene's impact on fitness. Indeed, the concept of "function" seems to naturally lend itself to being sub-divided into "essential" and "non-essential" sub-types. That's not to say that ENCODE didn't take this notion too far, only that "function" is not a self-evident concept (as I have already said) David https://www.blogger.com/profile/17990525425532211865noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-55983560657349019492017-07-31T15:17:30.391-04:002017-07-31T15:17:30.391-04:00It's very difficult to understand exactly what...It's very difficult to understand exactly what the authors did. As far as I can tell, they only analyzed a subset of clones - the ones that were not highly deleterious after induction. Among this subset of 1,082 clones about 70% showed an effect on growth of the population after induction. <br /><br />Keep in mind that following induction there will be hundreds of copies of the transcripts in the bacterial cells. That's a lot of copies. <br /><br />Among the clones that showed an effect there were far more that were deleterious than were beneficial. The breakdown was between 20-25% beneficial. If you include the 30% that showed no effect then there were 14-17% beneficial.<br /><br />However, the authors remind us that all the severely deleterious clones "are already mostly lost" as a consequence of the experimental design. Thus, of the total number of random transcripts only a small percentage (~1-5% ?) were beneficial under conditions where a huge amount of transcription was present. Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-42240096862197770152017-07-31T11:03:40.370-04:002017-07-31T11:03:40.370-04:00Yes, I agree that the net effect would NOT be &quo...Yes, I agree that the net effect would NOT be "somewhat beneficial". But going back to thinking about the functionality of individual transcripts, this does suggest that for any given sequence that is transcribed, there might be a fairly high chance (25%) that it does something beneficial regardless of whether it goes on to be conserved or not.<br /><br />Note that this study did attempt to look at whether the functions were a result of the expressed transcripts or the translated peptides.<br /><br />In the abstract they write:<br /><br />"Testing of individual clones in competition assays confirms their activity and provides an indication that their activity could be exerted by either the transcribed RNA or the translated peptide."<br /><br />To test whether the functionality was coming about through the expressed transcripts or the translated peptides, they added a stop codon to the start of some of these random sequences to see what difference it would make:<br /><br />"Given that the bioactivity could be conveyed by either the transcribed RNA or the translated peptide, we produced versions of these clones harbouring a stop codon directly at the start of the random part of the sequence, that is, only the first four amino acids that are common among all clones would be translated. These mutated clones were also tested in pairwise competition assays with the empty vector. Only one of the clones (clone 600) showed a clear difference between the mutated and the non-mutated version (Fig. 5c), which would suggest that only this clone exerts its effect via the encoded peptide, whereas the two other clones might act through their RNA alone. To study this in more detail, we did an experiment with a direct competition of each clone with its stop codon counterpart, but with the same qualitative results"<br /><br />If about 76% of our DNA is transcribed (according to ENCODE) and most of those transcripts are junk but 25% of them confer some advantage to the organism either through peptides or simple RNA transcripts then that comes to about 19% of our genome altogether.<br /><br />In real animals though I imagine that the proportion of useful transcripts is much lower than 25% because these randomly inserted sequences were all given promoters that I assume were causing them to be transcribed in high concentrations. <br /><br />As Moran has pointed out many times before, most of our transcripts are transcribed at frequencies of less than one transcript per cell. This probably significantly affects whether they can be functional or not.Aceofspadeshttps://www.blogger.com/profile/09534611408824723712noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-7090104552775546852017-07-31T10:11:34.160-04:002017-07-31T10:11:34.160-04:00That's what natural selection does.
Duh. Now...<i>That's what natural selection does.</i><br /><br />Duh. Now tell us what "genetic load" is, and you're halfway to understanding why your post that I commented on is nonsense.judmarchttps://www.blogger.com/profile/03111006189037693272noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-30983643476674542052017-07-31T09:54:13.254-04:002017-07-31T09:54:13.254-04:00Always possible. But, following your logic, 52% ar...Always possible. But, following your logic, 52% are slightly deleterious, so the net effect is not "somewhat beneficial" at all. Note, however, that the said random sequences were embedded with "enough regulatory machinery to make sure they would be transcribed and translated by the host". What you're talking about here are random protein-coding genes, and it's the expressed proteins that are beneficial or otherwise. Junk DNA is not generally transcribed (except as noise) and certainly not translated, so the case doesn't arise.John Harshmanhttps://www.blogger.com/profile/04478895397136729867noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-29491723035290734852017-07-31T09:46:56.601-04:002017-07-31T09:46:56.601-04:00An experiment was conducted earlier this year to s...An experiment was conducted earlier this year to see if researchers could come up with some functional denovo genes from randomised DNA sequences. The result was that they were able to create hundreds of functional genes from random strings of DNA of length 150bp (this should put to bed any argument from ID advocates which states that new, useful genes cannot arise from junk DNA)<br /><br />In an article about the study, one of the researchers recounts:<br /><br />"During my early months in the Tautz lab, while still a Master’s student, I contemplated the possibility of doing an experiment that could support de novo evolution as a general process, and so I came up with a thought experiment. I would insert random sequences in living cells, together with enough regulatory machinery to make sure they would be transcribed and translated by the host. Then, I would wait until any of those would mutate enough to “acquire a function.” It occurred to me that starting with a sufficiently large pool of random sequences would reduce the waiting time, because some would exhibit some biochemical activity upon their introduction."<br /><br />https://natureecoevocommunity.nature.com/posts/16396-exploring-random-sequence-space-in-the-name-of-de-novo-genes<br /><br />The results were surprising - 25% of the random sequences they generated were beneficial to the bacteria that received them and 52% inhibited growth.<br /><br />Here is the paper: https://www.nature.com/articles/s41559-017-0127<br /><br />My question to Prof. Moran then is: If 25% of the transcripts from these random bits of DNA were able to promote growth in E-coli then wouldn't that imply that the same might be possible for us? <br /><br />Maybe 25% of our own transcripts which are not evolutionarily conserved are also be conferring some small transient advantage? It might not be that this advantage is strong enough to be selected for and that is why it is transient - a few tens of thousands of generations from now and perhaps most will be mutated out of existence but by then other new transcripts that are somewhat beneficial will have popped up in the mean time.Aceofspadeshttps://www.blogger.com/profile/09534611408824723712noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-20506339856290888902017-07-27T22:16:45.914-04:002017-07-27T22:16:45.914-04:00@Peter
It's abundantly clear that some DNA is...@Peter<br /><br />It's abundantly clear that some DNA is functional even though the specific sequence is irrelevant. That point is not controversial ... at least not among knowledgeable experts. Spacer DNA is a good example. <br /><br />Your example doesn't make any sense and I doubt very much whether that explanation will hold up in the future.Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-66179856348917480362017-07-27T21:24:10.319-04:002017-07-27T21:24:10.319-04:00Thank-you for all your helpful comments. I'm s...Thank-you for all your helpful comments. I'm sure the experts in genomes and evolution have never thought of them before now. I'll make sure to pass them along to all the ignorant population geneticists who have been studying the genetic load of populations for almost a century.Larry Moranhttps://www.blogger.com/profile/05756598746605455848noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-62541499683114980862017-07-27T17:09:51.846-04:002017-07-27T17:09:51.846-04:00Solovei et al provides a pretty compelling effect ...Solovei et al provides a pretty compelling effect that heterochromatin (i.e. mostly repetitive "junk" elements) has a causal-role function on visual acuity in nocturnal mammals. Mechanistically, what's going on is that the rod cells compartmentalise their heterochromatin towards the centre of the nucleus, where the increased density (heterochromatin is more compact) acts as a mini-lens focusing the light.<br />http://www.cell.com/cell/fulltext/S0092-8674(09)00137-8<br /><br />Of course, a causal-role function doesn't necessarily imply a selected-effect function. However, a new paper shows that in owl monkeys - a species that is transitioning to a more nocturnal lifestyle - there has been an accumulation of specific heterochromatic repeats, and that these are the genomic elements involved in the light-focusing role. So that looks very much like something that has been selected for in the evolution of this species.<br />https://academic.oup.com//gbe/article/4048064/Co-opted-megasatellite-DNA-drives-evolution-of<br /><br />Of course, this only applies to nocturnal animals and isn't a general answer, however I think it's a very clear demonstration that there can indeed be selection to increase DNA bulk in a sequence-indifferent manner.Peterhttps://www.blogger.com/profile/12559721137290332762noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-8358792006373125412017-07-27T12:05:25.999-04:002017-07-27T12:05:25.999-04:00One more thing, Larry. Many harmful mutations have...One more thing, Larry. Many harmful mutations have RECESSIVE phenotypes and their effects are masked by the other allele.Anonymoushttps://www.blogger.com/profile/03338277048053429980noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-57653369308614513432017-07-27T12:01:59.524-04:002017-07-27T12:01:59.524-04:00Larry posits an argument from ignorance and person...Larry posits an argument from ignorance and personal incredulity: "I can't believe that so much of the genome is functional..therefore it's junk." He also thinks that an inactive retrotransposon cannot be functional because it can't move about even though he knows they donate a regulatory sequence (as with Alus). Anonymoushttps://www.blogger.com/profile/03338277048053429980noreply@blogger.comtag:blogger.com,1999:blog-37148773.post-58370304867810181022017-07-27T11:53:24.703-04:002017-07-27T11:53:24.703-04:00What Larry doesn't seem to understand is that ...What Larry doesn't seem to understand is that much of the genome consists of functionally redundant copies of genes. As such, a deleterious mutation can be compensated for by another intact copy. So the genetic load can be tolerated by the buffering effect of gene duplicates. Anonymoushttps://www.blogger.com/profile/03338277048053429980noreply@blogger.com