Those things can, however, lead to DNA being less and less typeable? In other words, the DNA be at levels where it becomes more difficult to get an answer, if one can get an answer at all?
Isn't it correct that the difference between medical specimens that are subjected to PCR and forensic specimens is where the sample came from?
Now, you are familiar in this case of testimony that has described that following PCR several genetic markers were typed using this dot blot technique?
And in fact you've observed the use of that technique in all approximately thirty cases you've viewed, right?
The direct dot-blotting, yes. It is a little different, but the same--essentially the same.
In other words, whether direct or indirect, that is not a major difference or is that a significance difference in terms of the opinions you have offered over the last two days?
That dot-blotting in your laboratory includes dot-blotting following PCR on vaginal specimens, correct?
At one time we did look for an organism, human papilloma virus using that technology, but we don't at the moment.
Yes. Well, it depends. I mean, basically our system at the moment, the way we run dot-blots is the detection methodology is what is called chemiluminescence and it results in basically the release of light during the reaction and so it is more light developing a photograph, so what we really keep are what would be similar to the negatives.
As far as your own experience in criminal cases, that has included looking at photographs of dot-blots, correct?
And in fact would it be correct to say that the analyst who sees a dot-blot strip live is in a better position to see exactly the reactions that have occurred?
One of the limitations of this system is those dots fade fairly rapidly, so it is true that you can--the analyst might see something that would fade away and then when you photograph it, you would--the second--it wouldn't be present in the photo where it could have been there in the original.
Errors are made--we do everything we can to prevent errors and I can't say that we have never made an error. They are a fact of life.
Well, some of the techniques that you use or have used in your laboratory have a certain error rate, don't they?
As far as this HIV testing in your laboratory, have there been sort of two phrases using that HIV testing using PCR?
Okay. As far as HIV testing, how do you conduct that testing just in terms of the technology used?
At the present time for the purpose of screening donors, if that is what you are asking me, we use what is called a serological technique which is looking at antibodies as opposed to looking at DNA.
In broad terms would that be comparable, as far as the broad technology, to the serology testing conducted by the Los Angeles Police Department in this case?
As far as that testing, did that then lead to a different form or different method of testing for HIV in your laboratory?
Well, it didn't really lead to it. It is just that that particular way of looking for a virus is an indirect way of looking for the virus. It just simply says that if you are exposed to a virus, you make an antibody which is part of your recognizing that foreign organisms and responding to it and so you can look for that as a marker to indicate that perhaps you've been exposed to that virus, but it doesn't tell you really if you are--if the person is infectious, so the better way is to look directly at the nucleic acid which is the infectious part of the virus.
All right. So we can refer to one method as an indirect method and one method as a direct method?
As far as the incorrect method, what error rate did that method have, as used in your laboratory?
Well, any of these are basically tests that are clinical diagnostic kits and before those are released for that purpose in any laboratory the FDA goes through validation studies that define what is known as the specificity and sensitivity of a test kit and the specificity means how specific is it to what you are looking for. That is, do you only detect aids or maybe in certain clinical situations might you get a falsification that it is positive but it really isn't aids. And as far as sensitivity, it says how many people who are infected are truly identified as being infected, so in any one of the aspects of any test, any laboratory test, is that none of them are perfect and what you need to do is identify the parameters and the guidelines that allow you to wait to determine how much weight to put on that test. So we use that kit, and the error rate that was reported for that kit, I think the original kit was probably 98 percent specific and 98 or 99 percent sensitive.
Isn't it correct that the earlier method, the indirect method, had an error rate of between ten and fifteen percent?
Haven't you previously testified that the error rate for the earlier technique was between ten and fifteen percent?
The early kits on that had a higher false positive rate and it has to do with the fact that you are looking for a virus that is found very infrequently in the population, so it is called low incidence, and that creates a problem in terms of--of this false positive situation.
Umm, it--there is no FDA approved test for that, so there is nothing published in terms of what that rate is. I'm sure it will have an error rate.
Haven't you determined in your own lab that with regard to this second form of HIV testing that you encountered an error rate of between two and five percent?
Have you in fact encountered an error rate of between two and five percent for any of your testing in your laboratory that you use that is techniques used in case work?
I think there are--you can find blind trials, for instance, of PCR testing, where error rates for detecting viruses are--are that high perhaps.
They are used and what you need to know is what the error rate is. The important thing is knowing the error rate so that you can look at whatever the data is and properly evaluate how much weight to put onto it based upon what your probability of making a mistake is.
In other words, would an error rate of, let's say, one to two percent, that is a portion of what a doctor may use in counseling a patient about whether that patient has a disease or not?
They take that into consideration or should that take--should take that into consideration. If a test result doesn't seem to fit with their clinical impression, they would retest.
All right. Dr. Gerdes, I'm going to ask to shift your attention to the DQ-Alpha marker. I think we spoke a little bit about it earlier. It is your opinion, isn't it, that there is absolutely nothing scientifically suspect about the DQ-Alpha gene?
You've actually personally. That is. Your laboratory. Looked at the DQ-Alpha marker, I believe you described that yesterday?
Your laboratory does more work with this other marker closely related called DQ-beta, correct?
It is more polymorphic and it is more relevant to the reason we look for it, which is for matching in transplants.
In other words, for your purposes DQ-beta is more informative in your laboratory than DQ-Alpha because it gives you more information about what you are in particular looking for?
Your experience, as far as DQ-Alpha is concerned, comes primarily from cases that you've looked at in your role as a Defense consultant, correct?
And I'm referring to the kit manufactured by Roche or developed by Roche, manufactured--marketed by Perkin Elmer used by all three laboratories in this case, correct?
Now, you express the fact that you are familiar with the user guide; is that right?
And is that--I think we have brought this up before. Does it look like what I have entitled "Amplitype User Guide"?
I believe there are--in general. There are certain specific areas of it that I disagree with, but in general it is a good guide.
Are there any publications, to your knowledge, scientific publications, showing that the use of the Roche kit, DQ-Alpha kit, is unreliable?
There are no publications that would specifically conclude that it is totally unreliable. There are numbers of publications that discuss the contamination issues and other issues we brought up.
KEY QUOTEDo you in fact use a kit in your own laboratory developed and manufactured by Roche?
There are, in this user guide--and actually let's talk about two things. First the user guide itself, it contains protocols on how to conduct PCR DQ-Alpha typing, correct?
There is also a second item that comes with the kit, correct, and it is a different document from this user guide?
Well, that is--it comes with the kit and it is pretty much abbreviated version of this larger user guide. It also describes more succinctly how you would set up the test.
In other words, it is also a cookbook but it is kind of directly relevant to how to do each step?
It is your opinion, is it not, that the user guide and the package insert describe correct scientific procedures for this test?
I think errors are a fact of life. Everyone makes errors.
You have never used the kit, correct?
There are no publications that would specifically conclude that it is totally unreliable. There are numbers of publications that discuss the contamination issues and other issues we brought up.
It is sort of the cookbook, if you will.
The prominent consensus is it does cause aids.