I remember it very clearly because I had never heard the term codon before and I’m especially anal retentive when it comes to jargon (even if I’m occasionally sloppy when using it).
And yes, I know that there are no codons specific to any genome. Jesus, do you imagine I’m sitting here drooling on my keyboard?
And no, I do not know specifically which sequences they used for their probes. Once I learned what a codon was and that specific sequences would be selected for amplification, I just assumed that they got a hit on one of these when they shouldn’t have. I didn’t think much about it past that.
I can tell you though that everyone regarded the situation as being quite serious - so much so that that they offered to have me and the woman I had gone in with redo the test gratis. And you need to understand that the test was several hundred dollars for each of us at the time.
As a biologist (though not a molecular biologist) I am familiar with PCR but thanks for the tutorial. I see on a quick Google search that PCR is used to diagnose HIV very early so I stand corrected. I see that you know that codons are not specific to a genome yet you seem convinced that a codon from yeast caused a false positive. Can you possibly begin to explain how one codon, which is not even specific to any organism, could result in a false positive PCR test? With one codon of information you could not even determine if something is a virus, plant, animal, or extraterrestrial life form.
I’m not convinced. I keep saying 'I admit it’s speculation.’ There, I bolded that time. M’kay? I will admit that I’ve proceeded on that assumption given the fact that that PCR is pretty goddamned specific, my test was positive for one ‘whoziwhatsits’ while the gf’s wasn’t pre-supplement and both were negative post-supplement.
This was only over a period of maybe 2 weeks tops so there couldn’t have been many confounding factors. I was working at the time and had a pretty solid routine, normal diet, etc. I should also point out that the gf was shitting bricks during the interim on the off chance that there was something to this. On the other hand, I didn’t have the slightest doubt even though given some of the shit I’d done in the past I probably should have.
So what it really comes down to is the use of the word codon. Honestly, now that we’re discussing it in some detail, much beyond what would have ever occurred to me, I’m a little curious myself. Maybe if we had someone here who actually works with DNA analyzers, they could tell us. For example, let’s say you have a 30 nucleotide probe you’re using, what would that mean for you to have one codon match? That certainly doesn’t make any sense, so I see your point.
OK then, is it possible that one of the tests was exclusionary then such that it was supposed to be a full mismatch? I can’t imagine what that would be, but then a one codon hit might make sense, right?
And I just pulled that example out of my ass. So if we had someone here who actually DID know what they were talking about, maybe they could give us a broader perspective rather our just sitting here pissing at each other. Ooooo, there’s an idea.
I spent six years doing PCR-based diagnostic tests on human blood samples for a living. I can chime in and say that what you’re saying is nonsense. I have no opinion to offer on where in the chain of information something got garbled, but the very concept of an “HIV-specific codon” is meaningless. Every possible codon is found in every genome on the planet. It’s a statistical certainty, given that there are only 64 possible codons, and genomes contain on the order of billions of nucleotides.
There are certainly plenty of HIV-specific markers, or sequences, or even genes, but the word “codon” simply does not work here. Sorry.
Oh, and I’ve spent the last five years working on a PhD in molecular biology, if you are doubting that I know what I’m talking about.
OK. First, thanks for posting, but second, we’ve been over this a few times and no one is saying there is an HIV specific codon. What we ARE doing is trying to figure out what might have been meant.
deltasigma, it’s clear from the preceding discussion and your own admission that you don’t know very much about the technical aspects of this subject. (I’m also basing my assessment on my own knowledge of the subject as a professional biologist.) Whatever you understood at the time, either the person who informed you did not communicate the information correctly or you misunderstood it. I would suggest that you stop trying to argue the point.
I’m genuinely surprised there is still the impression that I am arguing anything other than the veracity of my recollection. But since I’m clearly in the minority and having one of my ‘bad communication’ days (yes, it’s a thing I’m sad to say), I’ll just say that I hope I wasn’t too harsh in some of comments and now see what I can do regarding a graceful exit.
I do diagnostic real-time PCRs for viruses as part of my job. Not HIV, but the target of the PCR doesn’t make a lot of difference for what we’re discussing here. That being said, without more specific information on the test that was used in your case, it is very hard to answer your question. As myself and others have said, the word “codon” really has no place here, but let’s just assume it was used (somewhat inappropriately) as a shorthand for “three bases”. With your test taking place in the early 1990s, I assume it was most likely a “conventional” PCR or RT-PCR (as far as I can see, with retroviruses it could go either way), and not a “real-time” PCR. So, the specificity of the test comes from the sequence of the two primers, each probably around 20 to 25 nucleotides in length. The primers would likely be placed in conserved regions of the HIV genome, that means a region that is the same or very similar in different strains of the virus. At the same time, the person who designed the test would try to make sure that the primer sequences only match HIV, and not much else. I can’t give you hard numbers off the top of my head, but for a carefully designed 25-nucleotide primer, it is quite possible that the exact sequence only occurs in HIV. Take this random example: GAACACTCGGGTGGTTCTTAATAAA. That’s a forward primer from a rinderpest virus real-time RT-PCR. If you plug it into the NIH’s Basic Local Alignment Search Tool (kind of like Google for nucleotide sequences), you will see that the only hits with an identity of 100% and a query coverage of 100% are indeed rinderpest virus genomes. That being said, as much as 68% of the primer appear to be identical to a sequence contained in, amongst others, the human genome. Most likely that’s not a problem for the diagnostic test, though. In a probe-based real-time PCR, additional specificity comes from the probe, another short piece of DNA that must match your target sequence. But even in a simple 2-primer PCR, for possibly misleading DNA amplification to occur, it’s not sufficient if one primer kind of matches, the other must bind, too - on the same piece of DNA and reasonably close to the first. There are many more things to consider (annealing temperatures, location of the misannealing within the primer etc.), but in short, 3 bases out of 25-ish matching really is insignificant.
Because they come from once living matter, virtually all foods contain DNA, and, of course, all our own cells contain it. If it is a carcinogen it is an ubiquitous one that we have been evolving to handle ever since life first began on Earth.
This a bit like asking “Is oxygen a deadly dangerous gas?” Well, yes it is; it can burn you right up, but you would be a damn fool if you tried to avoid taking a good lot of it in with every breath.
