(BY MR. BLASIER) Dr. Gerdes, where there is contamination, does it always show up on controls?

Not always.

And if you do have what might be contamination show up on a control, does that affect the reliability or protect the reliability of other tests done -- results of other tests done at the same time?

Yes.

Is it accurate the PCR tests done by LAPD in this case in particular were all done at the same time?

Yes.

Pretty much?

Well, two different days they did the PCR.

Okay.

Now, you indicated expressing your opinion about the reliability of results in this case, PCR results, that you reviewed a videotape that was prepared by the Police Department Scientific Investigation Division, correct?

Yes.

And that was a videotape that was prepared by them for their presentation in the criminal case to demonstrate the technique that they use to collect evidence in this case, correct?

That's correct.

And let's start that video. It's 955.

This video is Andrea Mazzola collecting samples as she collected them in this case?

As a demonstration, yes.

Can you tell what she's doing there?

Yeah. I believe she's --

All right. Stop right there. Stop it there.

Oops.

Can you back it up just a little bit. Little more. Stop there, please.

You see her right hand?

Yes.

And what is her right hand doing?

Well, her right hand is resting on the pavement there, so even though it's gloved, she has now transferred anything that was on the pavement at that exact spot onto her glove.

Let's go.

Continue.

Stop.

Stop, please.

You know, she just picked up the tweezers in that right hand, so now she transferred anything that was on the pavement onto the tweezers. And the tweezers, if you look closely, I think are held within the hand so the point of the tweezers are touching the glove, so now the tweezers are contaminated. At least anything that was on the pavement there that she touched the glove to is now on the tweezers.

Continue, please.

Now she picked up those same tweezers. Now if they had anything on them -- she's taking an evidence swatch now, now they transferred onto the evidence swatch.

Now she's collecting the blood. This is a substrate control she's doing first here.

Which is supposedly a clean area separated from the blood spot.

Now, this is the substrate alcohol that she --

Correct.

-- that we talked about?

Correct.

So it's an area adjacent to the blood, presumably shouldn't have anything there. And that's the control. Now she's placing it into a plastic bag.

And she's touching about everything with her right hand?

Well, yeah, her right hand and her left hand are touching now and --

The things that were on her right hand could now be on her left hand?

Correct.

You agree that she's probably not even aware --

I'm sure she's not aware.

Objection, calls to speculation.

Overruled.

I think you can see that there are some swatches that are on the ground there that -- when she shook that out, they're on the ground. So the technique they use to get those swatches is not very efficient, but I'm sure they make, you know --

(BY MR. BLASIER) Now is she getting another swatch?

Right. And she's moistening that now -- you know, that I don't -- I don't know if I saw that or not because you could touch the dropper to the swatch. I don't know.

Stop there, please.

(BY MR. BLASIER) Now, do you see the moisture that she put on the ground for the substrate control; what has happened to that?

Well, the moisture has sort of spread out so it's touching the blood stain.

And would you agree that that could cause any material that's in that moist area to now get into the blood?

Yes.

Go on, please.

Now she's dabbing the blood, and she -- you can see she's trying to get the blood to come up on the swatch. She's actually moving that into the substrate control area too, so -- there -- you definitely see it there.

And the way this technique is performed, you can see that the rubbing back and forth is definitely going to be -- you can almost see flakes on that tweezers there of blood that's now on the tweezers. So you definitely got some blood on that tweezers, forceps.

She's taking the wet swatch and putting it in a plastic bag?

Correct.

What affect does that have on the degradation of DNA?

Well, it accelerates it because it's a nice, close, warm environment. It won't dry in that bag, so if it's -- as long as there's moisture, bacteria can grow, so it encouraged degradation of the DNA by bacteria growth.

That was then put in the back of a crime scene truck and left it in the sun for a number of hours.

In this case, that's correct.

What kind of affect would that have?

It's like an incubator. It's like a perfect environment.

You know, she didn't -- she's wiping the tweezer, she's only using water. And as sensitive as this technique is, you can't wipe DNA off. You should use a separate sterile instrument.

Now, I guarantee you could possibly get DNA off those tweezers again from that blood stain that she just took. Even though she's doing a pretty thorough job of trying to wipe it off, she's just using water.

Now the right hand --

That's not going to clean DNA off.

The right hand's going on the ground there?

Right. Correct.

Now she touched her leg there with the glove. Now what's she doing?

Now she's wiping the area with the glove.

And you also notice she didn't change her gloves in between, though presumably she's going on to the next specimen. She didn't change her gloves in between. So all of those incidents where she possibly got something on the glove, it's because she hasn't changed her gloves. Now they're still on the glove.

Now she appears to be getting an envelope and writing something on it, correct?

Yes.

And --

She's writing with that right hand and now has something on it because of the -- or may have, so now she transferred whatever that DNA is onto her pen, presumably the same pen she's using for recording the evidence all the way through the entire -- all the evidence she's collecting today or during this particular setting.

Now she got her hand on the ground again I believe.

She's brushing the ground?

Yes. And --

There's the pen?

There's the pen again.

Then she has the pen and the cap in the other hand.

Now, the cap has gone into the crime scene box.

Into the crime scene box. Now whatever is on her hand got on the cap is now in the box.

And again, she's moving onto -- she's trying to wipe those tweezers off with just water.

Now she has the end of the tweezers in her right hand?

Yes, she does.

So everything that's been brushed around with her hand is on the tweezers, and she's now introducing that tweezer into the swatch.

The cap, you notice, is on the ground, so -- the cap for the swatches, so if there's something there where she put the cap, now it's on the cap.

Now what's she doing?

She's doing another substrate control.

You can turn it off now.

(BY MR. BLASIER) Now, did you also evaluate in assessing reliability of the PCR results in this case, the procedure used by Collin Yamauchi to process the samples he analyzed on June 14 and 15?

Yes.

Is there anything important about the order in which he processed the items?

Yes.

And tell us what that is?

Well, ideally, forensic laboratories are aware that it's -- the greatest risk of cross-contamination is when you do what I described in explaining how it happens, that is handle a sample with a lot of DNA, and then handle a sample after that that's an evidence item which may have a very small amount of DNA.

So you want to try and set up the order that you do things so that, number one, you never do reference items which are known exemplars or known blood specimens or from known individuals, you never want to do those at the same time and in the same place that you do evidence items.

They should be separated, well separated, so that there's no possibility of accidentally getting something on your glove and then walk it down and -- or getting it on the pen while you're working there and transferring these things.

They should be done totally separately and everything cleaned in between in terms of bleaching down and making sure that there's no possibility of that transfer.

Now, that was not done in the LAPD -- by the LAPD. They handled Mr. Simpson's reference vial as the very first thing they handled. Immediately after that, the blood of the -- of the glove was sampled, in the wrist area of the glove, specifically. And then immediately after that, the Bundy drops.

So the sequence of events is -- is a setup for the possibility of transfer between Mr. Simpson's reference vial to the glove to the Bundy blood drops.

Now, Mr. Yamauchi also testified, I think, that on the first day that he did tests, he tested 23 items at the same time?

That's correct.

Is there any -- did you find any fault with that?

Well, that is an awful lot of items if you think about how carefully that you need to do everything when you do a crime scene.

For instance, what Gary Sims would do, changing paper and washing down the bench in between everything, and changing gloves in between everything is an extremely laborious, tedious process where you consciously think about these things at every step.

It will take all day to do four or five samples for Gary Sims. As a typical thing, he just takes his time, he makes sure that he does one item and he cleans up, and he does another -- the next item, and he is very careful and meticulous.

23 items at one time, and getting it through the entire processing, not only the DNA extraction, but the amplification of the typing, in one day, is a tremendous number of samples.

So the record is clear, isn't it accurate that every single item of evidence that was sent to Department of Justice and Cellmark was collected by the Los Angeles Police Department, using the techniques we've seen --

That's --

-- and processed through LAPD?

Correct.

Thank you.

Now, what is an artifact?

An artifact is a defect in the typing system that -- the DNA system that you're using. And what it simply means is you may get -- as I'm sure you're aware, on these DQ Alpha strips there are dots and those dots reflect, presumably, the presence of DNA.

Well, there are certain situations where you can get weak signals on those dots, and it doesn't really reflect the fact that there was DNA in the original sample. It's a defect in the way the sample is washed, so that you can get what's called cross-hybridization, that is a weak signal.

That is because of the fact that if two different genes are very close, and you haven't washed it properly, you can have these weak signals that show up, and it's not because of the fact there was DNA there, it's because of the fact that there's a problem in the way the sample was washed.

Okay. And where that's the biggest problem is where the genes come closest together in terms of their sequence, their very close, and it takes very -- what's called stringent conditions to be able to differentiate those.

And sometimes I'll have a weak signal, a bleed through, if you will, and it's not because there's any DNA in that particular allele there, it's because of this artifact, the washing artifact. So it's called an artifact.

You need to be able to recognize that to say that's not real, that's just something -- a problem with this particular way in which we detect the DNA after we've amplified it.

Now, if the testing kit that is produced by Kirk and Zelmer (phonetic) is followed pursuant to the -- the book that tells you how to do the test, should you see cross-hybridization show up?

If you use DNA -- greater than 6 nanograms of DNA, that is a fairly large number of DNA, actually, you would see this type of cross-hybridization. You certainly should not see that frequently. It should be rare. And my experience in looking at other laboratories is it is rare, you don't see that often.

Is it an indication that the test has not been done properly?

Yes.

Can you always tell the difference between cross-hybridization and contamination?

No.

Now, let's look at civil board 1279.

(BY MR. BLASIER) Doctor, you recall this board from the criminal trial?

Yes.

And -- now, this indicates the DQ Alpha testing strips produced by the Department of Justice for LAPD item number 30, which was a Bronco console stain, item 31, which was in the same area from the Bronco, correct?

Correct.

And this third strip down here that says QC816, what is that?

That's quality control. And one of the controls we didn't talk about is called a positive control, and that's simply a known sample that is run to ensure the strips are -- are good, and that the appropriate type is being found on a known.

And so this is -- at the Department of Justice, what they have is individuals at the laboratory where they know what their type is; they do blood stains, those are run in concurrence with cases and -- so that you can look at the typing results on those knowns that presumably came from, or did come from single individuals, and then that gives you confidence that there weren't any artifacts in that typing run.

And you should also --

So that you're running --

You should only see the two alleles that belong to the person that made up the sample?

That's correct. Every individual only has 2 alleles, that's two numbers. If you find three, that means that's either a mixture, meaning there's more than two individuals there, or one of those artifacts occurred, or there's contamination.

Now, the bottom strip indicated positive control.

What is that?

This is a control that is incorporated in the kit itself, the company provides this, and it simply is another mechanism of just checking that the strips were made correctly, and that the appropriate typing results is being obtained. It's a 1.1,4 genotype DNA that's just sent to the laboratory and they're asked to type it as a control.

Now, is it accurate that all of these four testing strips are all done at the same time?

Yes.

And so that these two controls at the bottom were done at the same time as 30 and 31?

That's correct.

Now, the fact that there's an indication of a third allele -- fourth allele, I'm sorry, QC816, what should that have come back?

That's a 1.2, 1.2.

You actually should have seen one allele?

That's correct. That's the homozygote person.

The person who got the allele from both parents?

Uh-huh.

That strip actually showed something other than what it was supposed to show?

It showed a 1.1 and hint of a 1.3.

You've heard the terminology hints and traces and hint of traces?

Yes, that's a nomenclature that they use at the DOJ.

Okay.

And the fact that you see occasions of two alleles that aren't supposed to be there, is that an indication that's a problem with the test?

Yes. It indicates -- this obviously is a quality control. It should have come from a single individual. Even though the singles are weak -- singles are weak here, it indicates one of two things; either there's a contamination event here or the hybridization was performed in such a manner that there's cross-hybridization, that's an artifact, and because of the washing conditions in this particular batch, you've got these weak signals.

Now, in the positive control, what typing -- what alleles should you have seen on the positive control?

The 1.1, 4.

And the fact there's an indication of a 1.3 tells you what?

Again, either that positive control somehow got contaminated or there is a problem with the washing and -- in this particular run, and the 1.3, either the washing, or the development time, I should qualify that, but one or the other has created an artifact so that that 1.3 dot which shouldn't be visible at all is visible.

And could that be contaminated as well?

It could be.

Now, the development lengths, what does that mean?

Well, these strips, after you've hybridized the DNA to them, then you go through a color development reaction that allows these dots to become visible.

And it's very -- it's analogous to developing a photograph.

So if any of you have done photography, you know that you can overdevelop or underdevelop a photograph when you're in the -- doing that process.

You allow it to go for a certain amount of time until the dots become visible and then you stop it.

And if you allow it to go for a longer period of time, then the dots are going to become more intense. The longer you let it go, the darker they get.

And so that's critical in terms of the intensity that you'll observe on that strip in terms of how long you allow it to develop before you stop the development.

Now, the Department of Justice in interpreting item 39 interpreted this as --

31.

I'm sorry, 31, as being a 1.3 representing a real DNA, correct?

Correct.

And they ignored the extra alleles on the controls, did they not?

That's correct.

And in your opinion, given what showed up on the controls here, are the results for 30 and 31 reliable?

Now, what is the -- in calling this a real allele, and calling other ones not real alleles, is there a great deal of subjectivity that goes into this?

Yes. If it's taken in a scientific test, it should be totally objective. You should be able to look at these and read off the numbers and that's the result. But because of this artifact, now it means on certain alleles, especially the one alleles, now it becomes subjective.

You look at it and you say, well, it's kind of, there's maybe something there, but you know, I really don't think that's real in this case. Then the next time you look at it, you'll say, well, maybe that is real, and your decision becomes a subjective decision as to whether to count it as real or not.

So the discrimination ability of the test has been lost, it's been converted from a scientific test to a subjective test, that in the opinion of the analysis -- the analyst, this is the interpretation, as opposed to a strict scientific just read it out and this is what the results mean.

And certainly two experts ought to be able to look at the same data and come up with the same conclusion, shouldn't they?

Yes.

Let's look at 1281.

Now, you recall seeing this board?

I do.

And on this board we have four strips.

Item 52, which is a Bundy blood drop, correct?

Correct.

QC877, which is a control quality. That should have come back as a 1.1 and a 4?

No, that one is 3, 4.

3, 4, I got ya.

Then a positive control that was supposed to be a 1.1, 4, correct?

Correct.

And these three strips were all run at the same time, were they not?

They were.

On item 52, a Bundy blood drop, there is evidence of a dot at the 1.3, correct?

Yes.

And if that is actual DNA, that would be an indication of a component in the Bundy blood drop that was inconsistent with Mr. Simpson, correct?

That's correct.

And what did the Department of Justice in interpreting this, what did they decide about whether this allele was real or not?

They decided that was not real.

And comparing the two charts, what they decided was real and not real, is that a good example of the problems with the subjectivity of this kind of analysis?

I think it is, yes.

And the results on the Bundy blood drop, if that's real DNA, are consistent with some other contributor contributing to that stain, correct?

That's correct.

And 52 -- well, it's actually a fairly large stain compared to the other Bundy drops, was it not?

Yes, it is.

It was -- about how much DNA was in 52?

Well, if I remember correctly, it was around 200 nanograms of degraded DNA, but on slot blot and interpretation of the RFLP, around 25 nanograms of undegraded.

And the other Bundy drops had all the way down to 2 nanograms?

Yes, ranged from 1 to 4.

And I think you testified that one drop of blood should have about a thousand?

Correct.

Now, look at -- look at civil 1275.

You've seen that chart before?

I have, yes.

Now, would you agree that in forensics cases, the reference sample from the suspects and from the victims is critical?

Yes.

I mean it's kind of like the hub of a wheel from which everything else is compared to, correct?

That's correct.

And the reference samples should be done -- because of the way they're collected, certainly should be clean, and only have one person's DNA in it, correct?

That's correct.

All right.

Now, at the top of this chart we have the DQ Alpha types for Mr. Simpson and the two victims and the GC locus of the polymarker system.

Why don't you tell me quickly what that is?

The polymarker is a similar system to DQ Alpha. It looks at five other genes other than DQ Alpha, and as far as the format is as to how it's interpreted and read, it's the same kind of thing, you're looking at dots, but it's -- the GC is one of those five and it's just a second gene that you would have to look at that's in a similar way to DQ Alpha.

Okay.

And Mr. Simpson's GC type is a BC?

Correct.

And he's the only one that has a B allele in that locus?

That's correct.

Now, did you evaluate the typing sheets produced by LAPD on the reference samples that they typed?

Yes.

And -- incidentally, those again were done in the same runs as the evidence, were they not?

Yes, this is a second day now, and what happened here is, item 12, which is a Rockingham sample, was processed, and then after that Nicole Simpson's reference sample and Ron Goldman's reference sample. They were all processed together and in that order.

And item 12 is a drop that was found on the Rockingham foyer?

That's correct.

That had -- that had a lot of DNA in it?

Yes, it did.

Now, what result did Mr. Yamauchi get for Nicole Brown Simpson's reference sample that was processed in the same run after item 12?

He records a 1.1, 1.1.

You have examined that strip, have you not?

Yes.

Is there evidence of an additional allele in that?

It's very faint, but I think I can see there is definitely a 1.2 -- a hint of a 1.2.

The only potential source among the three people involved here of 1.2 is Mr. Simpson?

Of those three people, correct.

Certainly there would be a 1.2 in abundance in item 12?

That's correct.

Now, you've examined the strip produced by Mr. Yamauchi for Ronald Goldman, have you not?

Yes.

And what -- what alleles show up on that strip?

The 1.3 and a 4 and he records a faint 1.1 which is definitely visible.

You agree there's a 1.1 there?

Yes.

Did you see a 1.2?

No, you don't see -- you see that dot, but in this particular case it's because of, again, the way this typing is set up, the dot for the 1.2 is not specific only to 1.2. It's 1.3, 1.2 or 1.3 or 4. It's called masked because the 4 dot would also light that dot up. You also have to consider whether that -- when you have a 1 allele plus a 4, then you also have to consider the possibility that the 1.2 could be there. It's a possible allele.

And that's really a deficiency in the DQ Alpha kit?

It's a deficiency in the way the testing strip is set up.

It doesn't have a dot specific to the 1.2, you have to do that by deduction?

That's correct.

The 1.1 that shows up in Mr. Goldman's sample is consistent with the 1.1 in item 12, correct?

Correct.

And the 1 -- the possible 1.2 is also consistent with item number 12, right?

Correct.

And could not have come from Mr. Goldman.

Correct.

At the time this run was done, isn't it accurate that the substrate controls for item 12 and other items that were run in the same run, were run one after -- I mean the sample is run, and the substrate control was run as you described it should have been done, correct?

Yes, according to what they recorded, that was the case.

And for this run which was on the 15 --

15.

How many items were run?

19.

And isn't it accurate that none of the substrate controls showed any possible extraneous alleles?

That's correct.

And is that an example of how you can -- how you can get possible contamination or extraneous alleles that didn't show up on the substrate control?

I believe it is, yes.

Now, you've reviewed the typing sheet produced by Cellmark for the two victims' reference samples, correct?

That's correct.

And Cellmark, in processing Nicole Brown Simpson's, got the correct result, but in the -- in the polymarker system, there is evidence of a B allele in the GC locus, correct?

That's correct.

The only source of that among the three people is Mr. Simpson, correct?

That's correct.

And in their typing of Mr. Goldman, they got that one right?

Yeah. In both of those cases, the C dot was extremely faint, so the strips were not developed nearly as long as they would have been -- as they were developed at LAPD. By looking at the C dot comparisons, you can tell.

Let me ask you about that.

Cellmark uses a shorter development time than the Department of Justice, correct?

Correct.

And isn't it true that if you use a shorter development time, that they -- that may mask the presence of contamination, that doesn't show up, because you don't let the dots develop?

That's correct.

Okay.

Now, Department of Justice also tested both victims in December of '94, and you've reviewed those DQ Alpha strips?

I have, yes.

And isn't it true that the Department of Justice, when they tested Nicole Brown Simpson's, found a trace of a 1.2 allele?

That's what they record on their sheet. That's correct.

This is not a masked allele because you don't have the 4 allele, you can tell --

That's a true 1.2.

True 1.2. Again, the only source of that among the three people is Mr. Simpson?

That's correct.

And for Mr. Goldman they also found a faint trace of a 1.1 and a possible 1.2, and this one might be masked because you have the 1.2?

That's right. You have to consider it because of the way the test is set up, but the 1.1 is definitely visible.

They recorded it?

Yes, they record it on their sheet.

The only source of that is Mr. Simpson?

That's correct.

Now, it's true that some of these might be cross-hybridization as well, correct?

Yes. Well, basically, the alleles that are showing up here are the alleles that can cause artifact or be due to artifact, but the interpretation that -- my interpretation of this is either there is a coincidental artifact that happens in all three labs, to come up with those same results, which is unlikely, I think, or cross-contamination happened between these samples and the cross-contamination occurred at LAPD and then was repeated -- retyped at Cellmark and DOJ.

Now, would you agree that the most reasonable explanation for that is cross-contamination, not cross-hybridization?

That's my opinion, yes.

In a reference sample, again, that's the known sample taken in this case from Mr. Simpson, and the two bodies at the autopsy, correct?

That's correct. So cross-contamination in this case means that a substantial amount of DNA had to have been transferred because all of those have a lot of DNA.

Incidentally, the B allele, there's nothing indicated in the literature that that could be caused by cross-hybridization, is there?

No. So that that is a second gene system. That confirms -- to my mind, confirms the DQ Alpha system in terms of an interpretation of this being cross-hybridization by a second gene system.

And these are all high-concentration DNA samples, correct?

Correct.

That's all I have.