I've always said that creationists don't ignore the evidence. Take fossils, for example. Both creationists and evolutionists have different ideas about how and when most fossils formed but the fossils themselves are real. When we say the fossils aren't evidence for evolution, we're not “ignoring” the fossils – we're disagreeing with certain conclusions about the fossils made by secular scientists.
What, then, am I supposed to do with the oft-repeated claim that human and chimp DNA is 98% similar? Actually, I've heard various estimates from 95-99% (which should have been a clue that comparisons aren't necessarily objective) but DNA is something we can examine here and now. It's not a conclusion we have to draw about something that happened in the distant past; it's an observation of something that exists in the present. I trusted the secular scientists that the similarity was there. I wasn't going to “ignore” the evidence like we're often accused of doing. I simply disagreed that it was evidence that we are related to chimps.
I can see that humans and chimps have certain, physical similarities and, if DNA works like a blueprint for building an organism, then creatures that are the most similar will necessarily have the most similar DNA. Creation would predict that our DNA is most like a chimp's, less like a bear's, and least like a bird's. That's exactly what we find so the high similarity between human and chimp DNA was never a problem for creation. I've come to realize, though, that my confidence in the reports by secular scientists about the DNA similarities was misplaced. It seems their bias toward evolution causes them to engage in some “monkey business” when comparing our genomes (pardon the pun).
The first red flag for me should have been the fact that chimp DNA is 8% longer than human DNA. Now, I'm not a math whiz or anything but I do know that if you have 2 strings, and one of them is 8% longer than the other, then they could – at most – only be 92% similar. That's assuming, of course, that the strings are identical in every other way but their length. In the case of human/chimp DNA, there are more differences than just their length.
The information in DNA is represented by four letters. If you could lay a human and chimp genome side by side, it would be difficult to compare them due to their different lengths. What scientists have to do is line up the parts of the genome that are most similar. Even then, there are still parts that are dissimilar. Where there are letters in one genome not represented on the other, scientists assume this information was either inserted or deleted from the DNA – they call these “indels.” This chart shows how the genomes are lined up for comparison.
In this representative chart, only the “alignment region” is compared – the DNA before and after the alignment region is ignored. Even the indels inside the alignment region are ignored. The highly touted 96%+ similarity in human/chimp DNA is essentially achieved by comparing only the most similar parts of the DNA and ignoring everything else!
When you consider the additional length of chimp DNA and factor in the indels within compared regions, chimp DNA cannot possibly be more than 87% similar to human DNA. If you do a letter by letter comparison of the two, estimates of their similarity range between only 70-81%.
So what does this mean to evolutionists? Not much, I'm sure. They're still going to trumpet an absurdly high percentage as evidence for their theory. But it's not evidence of anything because it's not even true. And if similar DNA is evidence for evolution, then dissimilar DNA should be evidence against evolution.
I don't ignore scientific evidence. I was wrong, though, to believe the highly inflated similarity between human and chimp DNA offered by secular scientists. I know better now. Secular scientists, on the other hand, are ignoring the evidence. They're lying about the similarity between human and chimp DNA.
The earliest measures of human-chimp genetic sequence similarity, back in the late 1980s if I recall correctly, did not involve direct comparisons of sequences. They used DNA hybridization: mix fragments of separated strands of DNA from two species, let them unite with strand fragments from the other species, and then see how much heat it took to break them apart again, and compare it to how much heat it took to break apart strands of DNA from a single individual. This didn't require actually knowing the sequence of nucleotide base pairs in either genome (though it did require knowing how much difference in base-pair sequence corresponded to one degree of temperature difference, and apparently not all researchers agreed on exactly how much this was). Note that there's very little room for subjectivity in this, though there was room for imprecise knowledge.
ReplyDeleteNow, when scientists moved on to direct comparisons of sequences (after these were discovered), there was room for subjectivity. If, e.g. you come across a copy of an endogenous retrovirus at a certain locus in species A that is not present at that locus in species B, do you count every base-pair in the ERV as a separate difference, or do you count the entire ERV as one difference? After all, it presumably ended up at that locus as the result of a single mutation (retroviral insertion is a known type of mutation), so why not count it as "one" difference? Or, part of the reason chimps have a larger genome than humans is that they have five copies of a particular gene that exists in only one copy in humans. Do we count every one of the thousands of base pairs in each duplicate copy as one difference, or do we count each gene as one difference (again, gene duplication is a known type of mutation), and then count as additional mutations only those places where the duplicate genes differ from our single copy?
Different methods yield different measurements -- but presumably, if you're arguing over whether the difference is too great to be spanned by six million years of mutation, selection, and drift, you'd be most interested in the minimum number of mutations needed to produce the difference in base-pair sequences, not just in the number of different base-pairs. In any case, there's no need to count those 240 million extra base-pairs in chimps as requiring 240 million extra mutations to produce.
In this representative chart, only the “alignment region” is compared – the DNA before and after the alignment region is ignored. Even the indels inside the alignment region are ignored.
ReplyDeleteI think you misinterpret your illustration. There's no indication that the indels are not being counted; rather, two different types of difference are tagged, one as "substitutions" (single nucleotide substitution, the most common type of mutation), while others are tagged as inserted or deleted regions (note that each indel can probably be counted as a single mutation -- or as multiple base-pair differences, depending on taste). And there's no indication that areas outside the alignment region aren't being counted. The problem is that chromosomes can be rearranged, and sections of the genome transposed. It's a bit like pulling a page of writing out of a loose-leaf notebook and putting it back in at a different place in the notebook. If you're comparing this copy of the notebook with another, you need to match the corresponding pages (the "alignment regions"), which might not be in the same order in different copies (so that you can't just pick a "start" of the genome and treat the entire thing as one huge alignment region).
Note that by any measure, humans are more similar to chimpanzees than horses are to zebras, and most young-earth creationists think that the latter two species are descended from a single pair of ur-equines aboard Noah's Ark ca. 5000 years ago. Also note that horses and zebras differ more in chromosome count (64 vs. 32) than humans and chimps do (46 vs. 48).
if DNA works like a blueprint in building an organism, then creatures that are the most similar will necessarily have the most similar DNA.
It's more like a recipe than a blueprint. It's also in a "degenerate code" -- that is, the genetic code has two to six different three-nucleotide codons for each amino acid used in proteins (plus three "stop" codons). So it would appear that in principle, human and chimpanzee cytochrome-c (to take an enzyme at random) could be identical yet coded for by a different codon for every amino acid in the enzyme. In fact, only one codon is different. And this is assuming that there is only one combination of enzymes that can produce a given structure, which is probably not true.
Steven J,
ReplyDeleteAccording to Nature, “The current [human] genome sequence (Build 35) contains 2.85 billion nucleotides interrupted by only 341 gaps.”
http://www.nature.com/nature/journal/v431/n7011/full/nature03001.html
Also according to Nature, “Best reciprocal nucleotide-level alignments of the chimpanzee and human genomes cover ~2.4 gigabases (Gb) of high-quality sequence, including 89 Mb from chromosome X and 7.5 Mb from chromosome Y.”
http://www.nature.com/nature/journal/v437/n7055/full/nature04072.html
So, according to Nature magazine, a “best reciprocal nucleotide-level alignment” of chimp and human DNA paired only 2.4 gigabases. The human genome, according to Nature, contains 2.85 nucleotides. Why don't you do the math. 2.4/2.85 = only 84.21% similar. How then can they conclude our DNA is 95.8% similar to chimp DNA? It's not a mystery when you read their method:
“BLASTZ was used to align non-repetitive chimpanzee regions against repeat-masked human sequence. BLAT was subsequently used to align the more repetitive regions. The combined alignments were chained and only best reciprocal alignments were retained for further analysis.”
I see. When you analyze only the “best reciprocal alignments,” is it really a surprise to find similarity?
Like I said, the chart is only representative. I don't mean it to be an actual graph of human or chimp DNA. The DNA molecule is considerably longer than the 57 letters appearing on the chart. It shows how, if you only compare the alignment region and ignore indels, the two strings are 94.8% similar but if you compare the entire string, they are only 64.9% similar.
Getting back to the Nature article, total genome is 84.21% similar but their conclusion is that it is 95.8% similar. It's funny how you accuse me of misrepresenting the information.
Thank you for your comments. God bless!!
RKBentley
I think you are conflating limits in the ability to compare sections of the genome with deliberately cherry-picking the data. It's rather like assuming that if a polling company only asks 3000 people whether they'd vote for Trump for President, then obviously 99.999% of the country wouldn't vote for him. Sampling doesn't work that way. They've compared 84% of the genome and found various similarities and differences; that doesn't tell us what the similarity and differences are in the sections they have not compared -- but "about the same" would be the most obvious first guess. These are reports on a work in progress, but I would not expect the final results to drastically differ from the initial results from DNA hybridization done thirty years ago.
ReplyDeleteI also think that "best reciprocal alignments" means, not "the parts of the two genomes that are most similar," but "the places where we're surest that we're comparing corresponding regions" (if, e.g. you were comparing two copies of the Bible for similarities and differences, you really wouldn't want to compare, e.g. the beatitudes in Matthew to those of Luke and count those differences; that wouldn't be a "best reciprocal alignment"). The uncompared regions are replete with myriads of repeating "junk," such as particular endogenous retroviral sequences repeated over and over. Such regions exist in both species' genomes and are hard to match up well. Note that ERVs can be duplicated and moved during cell replication, so that these regions can be pretty scrambled by a relatively small number of fixed mutations.
Your quotes do not mean what you think they mean. I am not accusing you of misrepresenting; I am accusing you of misunderstanding.
Steven J,
ReplyDeleteIf you're going to use scientific polling as an analogy then let me suggest how it's being used in this case: It would be like polling the members of a Democrat PAC in Nevada and concluding that 95% of the voters in the country would vote for Harry Reid as President. The part you're sampling is biased toward the result you're looking for.
Also, I disagree with your characterization of the method described in Nature. It says, “The combined alignments WERE CHAINED and only best reciprocal alignments were retained for further analysis.” In other words, “we compared only those regions where nucleotides would already join together.”
Finally, I need to remind you again that the chimp genome is 8% longer than the human genome (BTW, other sources say it up to 12% longer). If one genome is 8% longer than the other, how can they possibly be 95.8% similar?!! Do I misunderstand that a sentence containing 93 letters can in no way be 95.8% identical to a sentence containing 100 letters?!
You can white wash it any way you'd like. Evolutionists are intentionally starting with the regions that are already the most similar and saying they are VERY similar. Well, duh!!
Thanks for your comments. God bless!!
RKBentley
Do not insult me and then thank me.
ReplyDeleteI am not "whitewashing," I am explaining. And I answered the question that appears to puzzle you:
If one genome is 8% longer than the other, how can they possibly be 95.8% similar?!!
Because they are counting by the number of mutations that would be necessary to turn one genome into another, rather than by the number of individual base pairs in the genome. A single retroviral insertion, a single gene duplication -- or, potentially, the duplication of several genes or retroviruses all nestled next to one another in a single section of the genome -- counts as one mutation, just as a single nucleotide substitution (e.g. an "A" replacing a "C", etc.) counts as one mutation.
Come, the young-earth creationists whom you trust so much assure me that gene duplication cannot "add information" to the genome. True, this relies on the creationist definition of "information" as "that which does not correspond to any measurable phenomenon, but which we know cannot be increased by natural causes, but only by magic," but if you think they're right, it seems to me that you ought not assume that, e.g. those four extra copies of a particular gene that chimpanzees have count as a "difference" between their genome and ours (by more reductionistic and materialistic methods of counting, they count for at least four differences -- but not for much more than four).
I disagree with your characterization of the method described in Nature. It says, “The combined alignments WERE CHAINED and only best reciprocal alignments were retained for further analysis.” In other words, “we compared only those regions where nucleotides would already join together.”
Alignment has nothing to do with joining nucleotides together; the methods discussed in the Nature article are not the DNA hybridization methods (which did not require any direct knowledge of the nucleotide sequences) but directly compare known sequences (learned since the heyday of DNA hybridization).
Steven J,
ReplyDeleteJust a few quick points because I'm working on my next post already:
I did not mean to insult you but I still think you're trying to downplay what's going on. If you don't like the term, “whitewashing,” you can choose another but the net effect is the same. Scientists are only comparing the most similar regions and ignoring the differences.
To you last point, I did a quick search on the term “chained together” in the context of DNA comparison. It seems it describes an algorithm used when comparing large sections of the genome and not a literal joining together. My bad. My point is still the same, though. The researcher still only analyzed the parts that were already determined to be the most similar. Now, you've offered suggestions about retrovirus insertions or gene duplication that can cause large sequences of mutation in a single event. I ask you, how do you know what the differences were? The study says they weren't analyzed! Only the most similar alignments were retained for analysis.
Finally, Human and chimp DNA are being presented as highly similar. The impression is that they are nearly alike. Dawkins has said that DNA is like a language and human/chimp DNA is the same language, like English, only spoken with different accents. If scientists have a different meaning behind the absurdly high percentage, namely that it represents only the number of changes that have to occur to change one into another, they're being very coy about making their meaning clear. Why does everyone keep saying “similar” if it doesn't mean “similar”? If they mean what you claim they mean, then there is an extremely high degree of deceit (lying, fraud, con, or whatever term you prefer) occurring when presenting the “similarity” to the lay public.
Obviously I could not change only 5 letters in a 100 letter sentence and the result be a 20% difference but if what you say about what scientists mean is true, I'm even more alarmed by the oft-repeated, inflated degree of similarity. It's word-smithing on the level of Bill Clinton.
God bless!!
RKBentley