More Junk DNA junk from the DI

[Faizal_Ali]

@scd

so? they can pick up a group by random. the result should represent the rule if it was randomly.

The selection is NOT random, as has been explained to you. But, as usual, in one ear, out the other.

[scd]

@Faizal_Ali

The selection is NOT random,

what selection has to do with that? im talking about the author choice.

[John_Harshman]

@scd

thus removing the confounding factor of ploidy or the contaminating DNA of prey, which are likely to be the primary cause of the large genome sizes attributed to lungfish and amoeba, respectively

Almost certainly not true. That’s a excuse, nothing more. Anyway, their analysis focused on sequenced full genomes, which are biased in favor of smaller genomes.

[davecarlson]

@scd

“our analysis is focused on haploid genome composition, thus removing the confounding factor of ploidy or the contaminating DNA of prey, which are likely to be the primary cause of the large genome sizes attributed to lungfish and amoeba, respectively”

Lungfish have not experienced a recent whole genome duplication but do have lots of active transposable elements. In any case, their haploid genome size is still much larger than ours. See the following paper for details:

The Lungfish Transcriptome: A Glimpse into Molecular Evolution Events at the Transition from Water to Land

[scd]

@John_Harshman

Anyway, their analysis focused on sequenced full genomes, which are biased in favor of smaller genomes.

lets say that it is true. how it make any difference? remember that their study is about correlation between amount of junk and complexity. the result should represent a pattern no matter what the size of the genome is. if a genome is larger than other genome we should also find a larger part that is junk, even if both genomes are small.

[John_Harshman]

@scd

remember that their study is about correlation between amount of junk and genome size

No, that isn’t what their study is about. The correlation is a no-brainer. Their supposed point is a correlation between amount of junk and organismal complexity. And that’s biased by their sample. Humans happen to be among the larger sequenced genomes, just because of the sampling.

[scd]

@John_Harshman

And that’s biased by their sample.

how? if i take a bunch of small genomes and i check the proportion of junk and complexity. say that i find out that we see a good correlation between the junk and complexity (number of cell types). isnt it a good estimation?

[Faizal_Ali]

@scd

what selection has to do with that? im talking about the author choice.

So am I. The authors chose a minute and non-representative sample of genomes to compare. That makes their data useless.

[Faizal_Ali]

@scd

the result should represent a pattern no matter what the size of the genome is. if a genome is larger than other genome we should also find a larger part that is junk, even if both genomes are small.

Only if you assume the correlation exists. You cannot demonstrate the correlation by choosing a non-representative sample.

This is basic, high school level stuff.

[scd]

@Faizal_Ali

Only if you assume the correlation exists.

we dont need to assume anything. if we take say 100 (small) genomes and we find a correlation between amount of junk and complexity- we then have a good estimation in general.

[Mercer (John Mercer)]

@scd

Many scientists use those terms, including me. Scientists who know what they are doing to not equivocate between them. Do you understand?

of course that if we find a non coding region that do something we cant call it “junk” anymore.

Faizal_Ali:

Of course, we still call it non-coding.

So now that you’ve been straightened out on that false equivocation, it’s time to decide if your misrepresentation is willful or not.

[scd]

@Mercer

So now that you’ve been straightened out on that false equivocation, it’s time to decide if your misrepresentation is willful or not.

i think its because of my english.

[Faizal_Ali]

@scd

we dont need to assume anything. if we take say 100 (small) genomes and we find a correlation between amount of junk and complexity- we then have a good estimation in general.

Not if the reason you only have small genomes to work with also reduces the likelihood that your sample will include organisms with lots of junk in their genomes.

Organisms with small genomes are more likely to have had their genomes sequenced.

Organisms with lots of junk in their genomes are more likely to have large genomes than those without lots of junk.

Can you figure the rest out for yourself?

[Faizal_Ali]

@scd

i think its because of my english.

That doesn’t seem to impede your ability to understand the creationist crap you keep touting.

[Mercer (John Mercer)]

@scd

So now that you’ve been straightened out on that false equivocation, it’s time to decide if your misrepresentation is willful or not.

i think its because of my english.

I’m skeptical, since you edited your response from:

“isnt the term “junk DNA” was refer to non coding DNA in the first place?”

to:

“of course that if we find a non coding region that do something we cant call it “junk” anymore.”

The answer to the question you deleted is no.

So, do you now understand that the descriptions “noncoding” and “junk” do not mean, and never have meant, the same thing?

And that equivocating them is a standard IDcreationist deceptive trope?

[John_Harshman]

@scd

if i take a bunch of small genomes and i check the proportion of junk and complexity

The human genome is on the high end of sequenced genome sizes, but it’s not at all on the high end of all genome sizes. There’s your bias right there. Fugu was chosen for sequencing precisely because it has a small genome. Ditto for Arabidopsis, and so on. Thus the bias.

[Chris_Falter]

@scd

we dont need to assume anything. if we take say 100 (small) genomes and we find a correlation between amount of junk and complexity- we then have a good estimation in general.

Rule #1 of statistical analysis is that your sample must be representative over the population being studied. If it is not, then you cannot draw any conclusions about the underrepresented portions of the population under study.

You seem not to have had any awareness of this fundamental rule of statistical analysis. Now that you have been made aware of it, do you see why the unrepresentative sample does not permit Sanford to draw a conclusion over the entire biological domain?

[Faizal_Ali]

I know where I’m placing my bet…

[Art (Arthur Hunt)]

@scd

I am referring to the correlations using cell type numbers as a measure of complexity. You need to look at the last sheet of Supplemental Table 1 to get these values. This sheet clearly does not have 1700+ entries.

so are they basically lying or they just pick a random group out of all these species?

Neither, I think, The authors do mention that they only looked at a few species when they were considering cell type numbers. That may not have been very clear. As others have pointed out, they looked at a sampling, but it isn’t really proper to consider their sampling to be either random or representative of all species. It is difficult to tell just why the authors chose those species that are in the paper.

@scd

As an aside, @scd, I am curious about how you found this paper.

it was from a creationist source that i read a long time ago.

The source is pretty resourceful. As I said, this paper is tucked away in an obscure place.

[John_Harshman]

@Art

It is difficult to tell just why the authors chose those species that are in the paper.

It appears that they chose species with published genomes. There may have been further selection, but that’s the bias I noticed.