Unidentified DNA

[Skepticoid]

Please forgive me if this question has been addressed elsewhere in the forum, and if it has, if you would direct me to the thread(s), I would appreciate it. 

My questions are these: 

1.  What are the potential causes of determinant DNA tests to result in "unidentified," or "unknown?"  For example, someone is on the trail of the ever evasive "Bigfoot" and comes across a hair sample.  This sample is sent in for DNA analysis, and the results are returned as "unidentified," or "unknown."  What are potential causes of this result? 

I am familiar with the possibility that the DNA extracted could be damaged, and thereby rendering the sample unidentifiable.  I am also aware of the possibility of contamination of a sample. 

2.  Am I correct that these are two possible ways a sample could result as being "unidentified," or "unknown?" 

3.  Are there other reasons why a DNA result would return as "unidentified," or "unknown?"

I am trying to explain that "unidentified," and "unknown" DNA results do not automatically default to a, "because we don't know and it is unidentified, it must therefore be Bigfoot," and that this type of default is the basis of an argument from ignorance/incredulity. 

Thank you in advance for helping me clear up my grasp on this question.

[Bufofrog]

35 minutes ago, Skepticoid said:

What are the potential causes of determinant DNA tests to result in "unidentified," or "unknown?" 

Probably this is because you cannot get DNA from hair.  There needs to be a hair follicle attached to th hair to get DNA.  So hair submitted with no hair follicles will come back as unknown.

[Strange]

It may just be that the lab doing the analysis doesn't have a sample from whatever species the hair came from (wether dog, pig, human or something else). 

18 minutes ago, Bufofrog said:

Probably this is because you cannot get DNA from hair.  There needs to be a hair follicle attached to th hair to get DNA.  So hair submitted with no hair follicles will come back as unknown.

Wouldn't they say "no DNA found" in that case?

[John Cuthber]

Or if there was DNA, but degraded and/ or contaminated.

[CharonY]

1 hour ago, Bufofrog said:

Probably this is because you cannot get DNA from hair.  There needs to be a hair follicle attached to th hair to get DNA.  So hair submitted with no hair follicles will come back as unknown.

In addition to what others said, you can get  mitochondrial DNA from hair; it depends on what types of analyses you want to run..

 

1 hour ago, Skepticoid said:

What are the potential causes of determinant DNA tests to result in "unidentified," or "unknown?"

It depends on the type of assay. The vast majority of DNA testing is performed on diagnostic regions and comparison with reference data sets. But the actual loci as well as the database being used would define what you can or cannot identify. The description does not provide enough information. 

[Skepticoid]

Thank you.  I'm sorry, I should have been more specific in my question.  I realize that DNA does not come from hair, but from the follicle, and I should have been more specific.  Let us then assume that the hair(s) that one is submitting has an appropriate amount of follicular mass to extract DNA from it.  What then, aside from degradation or contamination (if anything), could possibly result in an "unknown," or "unidentified" result.  And...as I think about it, are "unknown" and "unidentified" the same thing in the result?  It occurs to me that they could have two different meanings of which I am not aware, given that my background is not in genetics.  Finally, what would one expect to see if for example that instead of being "unknown," or "unidentified," that the sequence is complete, yet that the species is not yet known?

Thank you CharonY for your response.  When you say my description does not provide enough information, what type of further information would be needed?  I guess, what I am trying to ask (and perhaps rather poorly, as my background is not in genetics), is what type of information do I need in order to have the description make sense? 

[CharonY]

17 minutes ago, Skepticoid said:

Thank you CharonY for your response.  When you say my description does not provide enough information, what type of further information would be needed?  I guess, what I am trying to ask (and perhaps rather poorly, as my background is not in genetics), is what type of information do I need in order to have the description make sense? 

Essentially there are a range of DNA tests in use. To test familial relationships for example one can amplify certain DNA regions of DNA and compare them to references e.g. from the parents. If you submitted a different species, you obviously would not get usable results. In other words, the type of DNA tests are highly specific, depending on purpose and will have to be designed differently for each application as well as species.

I should also note that the vast majority of these test are not based on sequencing, but mostly rely on a form of genotyping. I.e. one monitors genetic variation at a given chromosomal location.

[Skepticoid]

So, if I am understanding you correctly, let us say that we find a sample of hair, with follicles fully intact, in the Pacific North West of the United States.  That sample is then submitted for genetic testing for identification of species.  The DNA would be extracted from the sample, and the genotypes of this sample would be compared to other known genotypes specific to that geographic region.  The process would assess the genetic variation of specific chromosomal locations within the sample, and it would be compared to other known samples.

Am I understanding this correctly?

 

For the sake of curiosity, let us assume that this hair follicle sample that was submitted WAS from some new species.  What would one expect as a reported result?  What type of findings would be necessary to substantiate validation of a cryptoid? 

Again, I appreciate you taking time and educating me on this process.  I have a very limited scope of knowledge relative to this field.

 

[Ghideon]

4 minutes ago, Skepticoid said:

That sample is then submitted for genetic testing for identification of specie

Since there are different types of DNA tests, what type of test did the hypothetical lab use on the hyptetical hair sample? I guess the experted outcome could be method dependent.

[Skepticoid]

I don't know what types of tests would be used.  I am an absolute novice in this area, and am trying to wrap my head around what appears to be a rather complex process.  What types of tests would be standard? 

[Ghideon]

41 minutes ago, Skepticoid said:

What type of findings would be necessary to substantiate validation of a cryptoid? 

What is a cryptoid*? Is this about Cryptozoology?

 

*) Do you mean cryptid?

[Skepticoid]

Pardon my typo, but yes, a cryptid,  And, I suppose this subject could fall under the auspices of cryptozoology, but my interest is not in the cryptids themselves, but in the process of identifying an unknown from a sample. 

I am simply trying to clarify reports that are often seen in various television shows that the "DNA" they retrieved from a "sample" is reported as "unknown," or "unidentified," and I am trying to solidify my understanding that a sample that is reported as unknown or unidentified is not evidence of a cryptid, in the case of my example, a bigfoot.  I wanted to know, in layman's terms, the various reasons a result could be returned as "unknown," or "unidentified."

[CharonY]

1 hour ago, Skepticoid said:

That sample is then submitted for genetic testing for identification of species. 

So species identification is a very specific application and there have been techniques developed based on hair-snaring. A typical genetic target are conserved mitochondrial sequences, which can be obtained from hair samples even with few or no follicles attached (though yield is better if one gets more cells with them). In the past one would do restriction analyses (in short: amplify region with PCR, digest the amplified DNA sequence and look at the resulting pattern to compare with known species), though more commonly nowadays the amplified locus is sequenced and compared to a reference database (often also called DNA-barcoding).

 

1 hour ago, Skepticoid said:

For the sake of curiosity, let us assume that this hair follicle sample that was submitted WAS from some new species.  What would one expect as a reported result? 

Even if the sequence is not found, one can use conserved sequences to build a so-called phylogenetic tree. There, the sequence in question would be quantitatively compared to existing ones and based on similarities one could figure out how the unknown species is related to known ones (i.e. if it is a close relative, for example). As long as a conserved locus is used (such as e.g. cytochrome C oxidase you will always get relationship info (i.e. it would not return as unknown). One issue could be contamination by other species, but as long as you can get some clean reads out of the sample of interest, it is often possible to isolate the novel from contaminating ones.

 

Quote

television shows

That is a part of the problem, many TV shows, especially those who are more sensationalist usually do not care much about proper reporting. There is a decent chance that a) all they got is something mundane, such as regular wildlife and reporting it as unknown just sounds more exciting, b) they contaminated their samples and mostly got DNA from their producer who handled the samples inadequately but did not have the budget to repeat the whole thing or c) they got data, but did not bother to ask an expert to interpret it or d) they used a different method, but would require some elaboration. 

[Ghideon]

3 hours ago, Skepticoid said:

I am simply trying to clarify reports that are often seen in various television shows that the "DNA" they retrieved from a "sample" is reported as "unknown," or "unidentified," and I am trying to solidify my understanding that a sample that is reported as unknown or unidentified is not evidence of a cryptid, in the case of my example, a bigfoot. 

@CharonY have posted good answers, I'll just add a few layman notes. If the sample is indeed unknown or unidentified due to some reason that's all there is to say: "unknown" or "unidentified". Cryptids does not exist according to current evidence available in mainstream science. No lab is going to have big foot, chupacabra or other cryptids' DNA in store for reference. An unknown sample could possibly be classified as a potential new species, related to some other known species and call for further scientific research. But I can't find a scientific argument why "unknown" DNA alone would be evidence for existence a cryptid.

3 hours ago, Skepticoid said:

I wanted to know, in layman's terms, the various reasons a result could be returned as "unknown," or "unidentified."

Another related thread* was just posted, possibly an example where a DNA test could return "unknown". If a hair sample belonging to an extinct species was analysed, without any context or details known, the result might be "unknown"?

*) https://www.scienceforums.net/topic/120650-18000-year-old-dog-found-in-permafrost/?tab=comments#comment-1124191

 

[CharonY]

41 minutes ago, Ghideon said:

Another related thread* was just posted, possibly an example where a DNA test could return "unknown". If a hair sample belonging to an extinct species was analysed, without any context or details known, the result might be "unknown"?

As briefly mentioned, it all depends on what type of analysis is being conducted. If for example we use a species-specific marker that is not found in the species under investigation, we would not be able to amplify anything, resulting in basically no result. Similarly, DNA can degrade or be contaminated to such a degree that we will not obtain data, either. However, if the DNA is of sufficient quality to be amplified, either restriction analyses or sequencing will allow you to assess relationship to known species via phylogenetic reconstruction.

There is a paper by Sykes et al. (Proc Royal Soc. B, 2014 281:1789) in which they analysed hair samples which were attributed to e.g. Yetis or similar organisms. After thorough de-contamination they sequenced a part of the mitochondrial DNA and found that they were actually from a wide range of mammals. Samples attributed to bigfoot were found to be cow, coyote, deer, black bear, horse, racoon and human, respectively. Two samples matched a fossil record related to current polar bears. 

 

[Strange]

3 minutes ago, CharonY said:

However, if the DNA is of sufficient quality to be amplified, either restriction analyses or sequencing will allow you to assess relationship to known species via phylogenetic reconstruction.

Presumably this would require the lab to have access to a database of known species. Is that something that is readily accessible to most labs? Or just some more specialised researchers?

[mistermack]

i would think that that "unidentified" could be sometimes be the result of poor storage of the hair samples. A lot of these samples are historic, and have been kept as souvenirs for years, and possibly changed hands a few times, before somebody decides to get it tested. 

Bacterial action due to poor storage could contaminate and destroy what original dna existed. "Unknown" would be a surprising finding, it seems to say that the dna is there, but not of a known type. Maybe a poor choice of words, rather than an accurate description of the result. 

[CharonY]

5 hours ago, Strange said:

Presumably this would require the lab to have access to a database of known species. Is that something that is readily accessible to most labs? Or just some more specialised researchers?

Oh no, there are huge repositories that you can freely access so that is generally not an issue. Even if your specific species is missing you can at least take the closes assumed relatives and work from there. Also, when you sequence something new and publish it, you have to make to submit it to one of those public databases.

[nuDAN]

May I dig up this thread, please? 🙂 Though I am not a scientist,  I have been pursuing this Sasquatch subject from the genetic angle for some time now. Oh yes, I have questions. This is on the subject of the NOTCH2NL gene, or should I say, its genetic variations? The question itself is hypothetical. If I take an environmental DNA sample and the metabarcoded test results includes Human, could I then run an barcode assay designed to target the NOTCH2NL gene? Specifically targeting, the Human NOTCH2NLA, B, or C gene variations at their 1q21.1 locus? And if so, what would it mean if none of those Human variations were present but the result did show a NOTCH2NL pseudo gene along with a defective copy at the 1p12 locus?

 

i can give a bit more background on the question, but regardless of anything Sasquatch, I am looking at this as a purely scientific inquiry. Mostly to understand whether or not a lab can design an assay to target one, two, or more, of those Human specific gene variations. I'm not disappearing but it is late for me here so I'll check in again later. Thanks.

Edited February 13, 2021 by nuDAN

[nuDAN]

I would like to add that I am aware that in order to zero in on the loci of a chromosome, I would need whole cells that have intact nuclei. Obviously, the question is in lieu of simply counting the number of chromosome pairs. 

[CharonY]

Fundamentally you can target any locus you want. You basically just need the right primers (which in turn requires that you have somewhat conserved sequences that those primers can bind to) and amplify the region for sequencing (technically you can also do direct sequencing without amplification, but there is still a step requiring primers to build the library, but anyway...). Assuming you sequence a locus and find sequence variation, depending on divergence you could estimate whether the sequence belongs to a given species (even if the variant has not been identified before). 

However, in order to look for duplication events, usually other, more traditional methods can be more suitable. Sequencing is really more helpful when you look at SNPs or other minor sequence changes.

[nuDAN]

Hey, thanks for responding, CharonY.

1 hour ago, CharonY said:

Fundamentally you can target any locus you want.

Yes, correct, as you say, the hypothetical could be aimed at any number of loci as many are Human specific. There are other primate loci that are, say, Chimp, Bonobo, or Gorilla, specific as well. To tell you the truth, there is so much going on with this hypothetical, I didn't know where to post it. Unknown DNA seemed like as good a place as any.

I also agree that specific primers and assays, even if assays have to be tweaked depending on a certain geographical region due to something like soil composition,......I can easily paint myself into a corner if I write too much so I'll try to be brief. This is about Sasquatch existence vs. non existence. So much has been said about its advanced primate body coupled with a relatively primate, perhaps even ape-like brain. So targeting the NOTCH2NL gene is the focus since there are no NOTCH2NL ape genes in the Human genome. The Great Apes, minus the Orangutan, all have a NOTCH2NL defective copy. That copy got repaired at the Chimpanzee/Homo split. Humans now have the NOTCH2NL along with three more copies: A, B, and C. Two paper in May of 2018 attributed the larger brain and cognitive boost in Humans to there "newer" variations.

Hypothetical: When getting Human contaminated samples from e-DNA field samples in North America, I propose looking deeper into the samples to see if any of the Human a, B, or C variations are present, OR if what is found is the Great Ape's NOTCH2NL's defective copy with could indicate an ape brain somewhere in the wild. It's never seems easy to articulate this concept, but that's what this is about. Nearly every piece of DNA that was initially suspected as coming from a Sasquatch source has been labeled as simply Human contamination. And yet I see no one thinking of ways to target either Human specific or Great specific genes. I don't think it is something that casting a wide net using a metarcoding protocol would catch. If the Sasquatch is that genetically close to Humans then why not run a cytochrome c oxidase barcode? At least that might work if the Sasquatch and the Human is thought to be that closely related.

On the base pair front, Sapiens/Denisovans- about 400 base pair differences, Neanderthal- 200+, Chimpanzee- 1,400+. With what folks think regarding the Sasquatch, it might be somewhere around 800-1,000 base pairs different? This is a discussion I've been waiting to have for some time now. Academia, so far, has walked away after any initial correspondence. So who is there to run this by? That's why I'm here. It's about the NOTCH2NL and it variations with respect to using a COX1 barcode protocol and figuring out just what is needed to do that.....outside of getting a Sasquatch onto a slab, LOL!

Edited February 16, 2021 by nuDAN