(BY MR. BLASIER) You said, Mr. Sims -- at the break, we were talking about windows. Do you recall that?
And your lab uses a window to -- because of the limitations of the technology and the limitations of your ability to really measure these fragments precisely, correct?
That's the total.
In other words, if two samples were to be declared a match, they would have to be no more than a total of 3.6 percent apart, and that would define the extremes.
Okay.
So hypothetically, a band that is -- if we talk about a 10,000 base-pair band, you would declare that to match to a band that is as small as 10,000 minus 360, right?
Well, there's one technical glitch here, and that is that our match criteria only applied below 9,416 base pairs you're up at that high end where we don't have a match criterion. That's why we don't use those bands in the statistical calculations, either.
We visually look at them -- that's important -- that's the first part of the process, to decide visually if they match.
We would do the sizing after that. They can be more at that high end because we don't have the match criteria that's the same up there.
Under your match criteria, the lower end of what you would declare a match is 10,000 minus the 360 approximately, correct?
And I can't reach -- (indicating to handwritten diagram).
THE COURT REPORTER: Excuse me. What number is that?
(BY MR. BLASIER) So you would declare a match for any bands that your computer told you were within this range, correct, when you're comparing it to a band here?
When -- yes, when we're comparing it to a band there, at that -- at that particular point, then we could go 3.6 percent in either direction for that particular band.
And you would agree, would you not, that in reality, if there is a single base-pair difference between your -- this one and this one, it's from different people, right?
If -- Well, you would possibly see more like a 16 base pair, something like that.
It would be a different repeat. You know, there's a technical reason for that.
And Cellmark -- your testimony was approximately plus or minus 2.6 percent. We used 2.5 because the numbers are easier. It means 5 percent below and 5 percent above.
So you would agree, would you not, that under your window, if something fell outside of your range, you wouldn't call it -- you would say that it did not match; whereas, another lab with the same data might say it's a match?
It -- well, again, part of the assessment, though, too, is the visual assessment when we see something at that level, say 3.6 percent -- I mean, I would be very suspicious that those two samples would match, just because it's so rare that unless there's some real change in the sample, that you would see that much of a shift.
That's an incredible shift. And so that's why we do so many of these probes, is to see that they line up all the way down, across all the Autorads.
You would agree that you allow yourselves this much of a tolerance in your measurements, don't you?
We do on a single-band basis. And again, we would -- we would have to see that consistently through all the probes.
Okay. And a band -- two bands that your computer told you were two percent apart is treated just the same as -- just as good a match as two bands that your computer says are really the same, right?
Well, when you when you look at it on the -- on the basis of a single band, that's correct.
But part of this whole assessment is looking at the entire profile across all the Autorads. And I would never expect to see a change of that sort of magnitude in one direction and then a change in the other direction of that sort of magnitude. That wouldn't happen.
Okay.
But you -- tell me, the most -- the highest number -- highest RFLP number for a stain consistent with Mr. Simpson was what?
Do you remember?
Well, it could be that there's a single-band pattern in one of those, but I think -- I think Mr. Simpson was a double-band pattern in all those loci, I think. That's my recollection.
So of those bands, can you tell me how many did your computer tell you were actually the same?
Well, to save you a little bit of time, isn't it true that in most of these cases, the computer tells you that the bands are not the same?
And we were talking about the quality of the DNA or the quantity of DNA in that back gate stain, correct?
And you were asked questions about the fact that drops that are picked up off the concrete might have different quantities of DNA than blood that is on a surface like the gate, correct?
Okay. Thank you.
Yes, that is our number 44, LAPD number 51, blood stain collected from the front gate.
And that was the same kind of surface as 117 on the back gate that was collected three weeks later, correct?
I remember seeing some photographs and I think -- I think it looked somewhat similar. I don't remember if they're both white or what exactly -- if they're the same, but they're similar sort of metal surfaces, as I recall, painted metal surfaces.
Now, you did testing on item 51 from the front gate. And the DNA in that sample was severely degraded, was it not?
It was. Actually, the overall total quantity of human DNA was estimated to be about four times more on the gate, but the DNA on the -- on the front gate was degraded human DNA.
Okay.
Now -- and the amount of DNA on the rear gate, also, when you try to quantitate how much is there, again, these are your estimates, are they not?
Yes, these are estimates. Some of them are based on slot blots for a human probe; some of them are just based on total yields of DNA.
And the back gate stain had more DNA than the Bundy drops and the Rockingham drops, correct, in the Rockingham driveway?
-- on quantity. And the estimate at that time was that number 6 -- that 117 had four times as much DNA as number 6. Do you remember that?
Yes. I think when we did this, though, we did some kind of calculation where we looked at nanograms of DNA per milligram of swatch material, because all these samples were weighed.
But as far as the overall yield, the number 117 had about 1008 nanograms; whereas, the number 6, the Rockingham sample, had about 56 nanograms. So it was about twice as much on the rear gate as the Rockingham drop in terms of total DNA from all the swatches and that sort of thing.
That's one of the variations in these kinds of calculations, how these samples are collected and how concentrated the swatches are, how much variation there is among and within swatches.
And you calculate or calculated that item number 47, the first Bundy drop, that 117 had about 27 times as much DNA as 47, correct?
Yes. It was -- it was a -- it was about 4 nanograms that we got out of it, 4.3 nanograms. Some of the other ones had less.
Okay. Bundy drop 48 had -- 117 had 45 times as much DNA as Bundy drop 4, did it not, approximately?
Now, 52, you got more DNA out of 52 than in any of the other Bundy drops, correct?
You were able to do it?
Excuse me.
No. On our -- there was a little difference here because our sample of 52 was a very small, little piece of the swatch. So on the basis of what we got out of that one swatch that we tested, because there were two that came back to us, we tested one of those and got about 3.6 nanograms out of that.
I think, again, as we look at this on the basis of this nanograms per -- nanograms DNA per milligram of swatch, it's about -- that's about right, 11 times.
Yes, that's my understanding, that's the sample that Cellmark got the RFLP result on.
We tested it for just the PCR markers, DQ alpha and D1S80.
(BY MR. BLASIER) Now, item number 117, stain from the back gate, had a great deal of genetic information in it, didn't it?
You didn't -- in processing item 117, you did not do anything to that stain that would have added EDTA to it, did you?
I did nothing that would deliberately add it.
I mean, the samples -- we use EDTA as part of our chemicals in our laboratory on a routine basis. We weighed some of these swatches out on an analytical balance that was probably about, oh, three feet from a couple kilograms of EDTA, but we took measures to prevent any of that EDTA, certainly, from getting on any of those swatches.
So -- but when you do the actual extraction, then you do use EDTA as part of the chemicals. We left a portion of that sample by itself untested.
In terms of your using it in these kinds of tests, that was a change of protocol that happened after your work in this case, was it not, based on an article that came out?
No. I think you're confused here, because EDTA is a standard chemical used in any molecular biology laboratory.
It wouldn't be added to the swatches, but when you take the -- some of the swatches and do the extraction of the DNA, to get the DNA out of the blood stain, there's EDTA in the chemicals that you add to that blood stain. Okay.
Okay.
Now, I put up the Bronco automobile board, which is -- I'm not sure what number it is. I think it's on the back. I think it's 131.
One of your results from the center console, stain number 31, you indicated had a weak 4 and a very weak 1.3, correct?
I remember that they were both weak. I don't remember the averb "very" being put on the 1.3, but it is present on the exhibit there.
And even though they had different intensities, you concluded that that was consistent with Ronald Goldman, correct?
And when you get a sample from evidence, you're going to get an equal amount of one allele as the other allele, correct?
Well, one of the limitations of the test, for example, when you look at Mr. Goldman's reference blood, you don't see a perfect, absolute balance of those two alleles, and you can even see that in our graphs.
So they're balanced, what we call balanced, but they're not perfectly balanced. You might discern a slight difference, even though they're all the alleles from one individual.
MR. P. BAKER: The exhibit board is 293.
Well, yes. The dots respond in a balanced fashion, but it's not what I would call a perfect balance.
Let me show you 2185.
Let me show it to you in person, make it easier.
These are some testing strips from your lab run out samples, among others, in the Bronco?
(BY MR. BLASIER) Now, if we look at item 29, which is the stain from the steering wheel -- correct?
And we zoom in a little bit on it, you called this 4. That faint dot at 4, you called that as a real allele, correct?
And you also -- I think you testified on direct that you're not going to rule out that there might be a 1.3 there, as well.
Did you say that?
Excuse me. I don't see a 1.3 at all on the strip.
What I'm saying is that I couldn't absolutely rule out the possibility that that weaker DNA -- this is a mixture, right? This is a mixture. And I couldn't absolutely rule out the possibility that there was an allele from a second individual who has a 4 dot, but who also might have another dot that is just not showing up in that sample.
It's -- well, the point of that is that this is a very technical issue and it revolves around the fact that the main type is a 1.1, 1.2, and that there are similarities between those alleles with this 1.3 allele, such that when we do the PCR amplification, potentially we could lose the 1.3 compared to the 4.
What I'm saying is, I could not rule out the possibility that there was a weaker contribution; therefore a 1.3, comma 4 individual.
I certainly don't think there's anything in that sample that says, of course, that's what happened. I'm not saying that at all. Because, for example, you could have a second individual who's a type 4, 4 or a second individual who is a 1.1, 4 or a 1.2, 4. There's a lot of possibilities there.
So when I see this kind of sample, what it tells me to do is, let's look at some other markers, get some additional data.
Now, the dot here at number 4, is that a hint, a trace, weak, very weak? How do you describe it?
That's a weak dot.
But the other thing that's important when you're looking at these dots is, you look within one of these strips -- in other words, you study this strip horizontally, you don't go up to the one above it, you don't go down to the one below it. And if you look at this particular strip we are amplifying, I believe, in this case, in this sample, on the order of about 400 picograms, as I recall, of DNA, which is an extremely minute amount of DNA.
And it's reflected in the fact that the dot that's to the right of the letter C is also very weak. And so what we do on any given strip is, we score the dots in relationship to that C dot.
So in this case, that C dot was present. It was weak; we called the 4 dot at the same level as the C dot. Whereas, if you look at the right and see the 1.1 dot, you see that is greater than the C dot.
It's a real allele.
One of the things about this system is that, when you have a nominal, what we call a nominal dot is those dots to the left, that those are definitely real dots, and you can see them.
(BY MR. BLASIER) Now, on the Bronco (indicating to Exhibit 293) -- on the Bronco, all of the stains that have the lower numbers, you're aware, were collected on June 14, correct?
And the only sample that you say is consistent with any blood from either of the victims is number 31, correct?
And our number is 29, I believe.
Yes.
When that fiber was actually taken off the carpet, it was given the number 293, correct?
As I understand it, the carpeting material was collected on the 14th, and then later it's given a number of 293.
I don't -- I don't know exactly the history there.
Okay.
The actual sample here on the carpet that the fibers -- that were taken off the carpet and then tested, that wasn't done until much later than June 4; is that your understanding?
All of the -- all of the other ones with the low numbers were done on June -- collected on June 14, correct?
And the only one of those that indicates the contents with either of the victims is number 31, correct?
And a positive control is known DNA, you know, what the type is going in, and you hope to see the same type on your testing strip, right?
Well, the tests are designed so that the positive controls come out the same type that they are known to be, correct?
Now, cross-hybridization is something that happens when you put too much DNA in the sample, correct?
That's one of the possibilities.
Also, if hybridization conditions are such that they're a little bit off, you could get some cross-hybridization from that, too.
When you say "a little bit off," you mean a little different than what the manual says to do?
Well, they may be off by, say, a degree or something like that in temperature, something like that, but just slightly, slightly off.
So the test is designed, if you do it precisely the way the manual says to do it, you should not get cross-hybridization, correct?
No, I'd say that's not correct. It is a matter of degree.
In our laboratory, you will sometimes see traces of cross-hybridization, and you can still understand that as being cross-hybridization; it doesn't change the types.
You always have to be concerned with contamination. That's why you run a lot of these negative strips that have no DNA.
Okay.
When you ran the sample for items 30 and 31 from the Bronco console, you ran a positive control, did you not?
Well, excuse me. I believe these were actually run by Renee Montgomery when the typing was done on the 30 or 31.
And this is Criminal 1279.
The above control in that run showed up a very faint 1.3 dot, did it not?
Yes. I'm sorry. I'm looking at the right page now.
There was -- there was what was called a hint of that 1.3 dot there.
That's one of the shortcomings of this test, trying to make these kinds of assessments of whether there's even a dot there, isn't it?
Well, the shortcoming is when one tries to overinterpret these results, clearly one can see what the dot pattern is.
I would say cross-hybridization is most likely due to the fact that the other controls, the negative controls, are all negative.
A quality control sample. We have QC samples. These are samples that the analyst takes and processes at the same time as the evidence samples are being processed.
And those samples are blind to the analyst. The analyst doesn't know what the correct results are.
Well, we did something like 20-some of those in this particular case, and we got them all right.
There can be some cross-hybridization, but it's more likely to be, more like what we call DX amplification.
You have to be concerned with that; you have to realize that this can show up in these typing results; that's why you don't want to interpret those very faint dots.
Whereas, the 1.3 here and here, you said we're going -- those were okay; that's not DNA; we pass the test, right?
I don't think so. You said he passed the test. He's a trying to explain to you what he passed.
Even though you got hints or traces or whatever you want to call them, of alleles that shouldn't be there, correct?
Okay.
Now, development time out here.
This figure represents the lengths of time that you allowed these strips to soak in the substance that has the DNA, correct, to develop the dots?
Well it's a color development. At that point, the DNA is -- has already come and gone, at that point there. Then there's a color development phase that goes on.
And in fact, the lengths of time that you use to develop these things can affect how intense the dots are, can't they?
Well, once you get to about 20 minutes, there's not much change after that.
I mean there -- most of this result takes place within the first five or ten minutes. But in our laboratory, we let them go from 20 to 30 minutes.
There are differences, are there not, with different development treatment times in terms of the intensity of the dots?
Isn't there?
I think once you get out to 25 minutes, you're very close to a plateau that's not going to change significantly.
There was a paper that came out not too long ago that showed that if you let these develop more than 20 minutes, you could have some dots that actually disappear, correct? Particularly from the polymarker system?
Do you recall a change in protocol being made in the DQ alpha, as well as the polymarker testing, because of that phenomenon?
(BY MR. BLASIER) This is board number 1281. And these are testing strips for Bundy drop number 52, correct?
Okay.
Now, 52 is a Bundy blood drop that you have said is consistent with O.J. Simpson and only O.J. Simpson, correct?
Well our -- our testing was only DQ alpha and D1S80. It was the RFLP testing was done and Cellmark.
Your testing -- your conclusion was that it's consistent with O.J. Simpson and not either victim, or not some third party with a different type, correct?
You call that as a 1.1, 1.2?
I think -- I thought I saw something on that one that I called -- I guess you say it was very faint traces in the wording, but there's barely something there, yes.
Mr. Sims, isn't it accurate that in deciding these faint dots, you interpret them in a way that helps the side that you're working for?
KEY QUOTESo, picking up the other board, board 1279, the 1.3 dot that you said was real and the one on the other board which is not real, they are very similar in intensity, are they not?
Thank you.
Now, you also, as you said before, did your own typing on the victims' reference samples, did you not?
And again, this was from the -- from the cards, not from -- you weren't sent the reference file, correct?
Looking at board 1275, isn't it true, Mr. Sims, that you found in Nicole Brown Simpson's reference sample, a possible 1.3 allele and a possible 1.2 allele?
I'll have to look at my notes on that.
I identified a trace in the 1.2 dot and what I call a faint trace, which is a weaker result, in the 1.3.
And one Goldman's sample, you also found an indication of a possible 1.2 and a 1.3 -- I'm sorry -- and a 1.3?
The only source among these three people of a 1.1 and a 1.2 in Mr. Goldman's reference sample is O.J. Simpson, correct?
I'm looking at the photograph.
And clearly, I -- I -- perhaps we should pass this around to the jury -- but these dots are extremely faint, and I don't think they're necessarily representative of any true alleles showing up.
These look like the types that you see from cross-hybridization or the DX phenomenon that I mentioned.
Isn't it accurate that in deciding these faint dots, you interpret them in a way that helps the side that you're working for?
I would say that they are similar, yes.
I doubt if any of them were exactly the same.
It had about twice as much as one of the Rockingham drops... the number 117 had about 1008 nanograms; whereas, the number 6, the Rockingham sample, had about 56 nanograms.
I'm sorry. We'll call these 'not real,' right?