Dr. Weir, with regard to your describing the approximate frequencies for mixtures that you just spoke about with Mr. Neufeld a few moments ago--
--what's your opinion about whether or not such frequencies should be placed on mixed samples?
I think they're totally misleading. They completely ignore the fact that there is a mixture. We are trying to interpret the evidence of a matching stain in a mixture. Not to take into account that it is a mixture I think is misleading.
KEY QUOTEIn particular--and let's set aside for the moment the way Mr. Neufeld asked you or had you calculate frequencies for some of these mixtures. Aside from that, what about applying the frequency to mixtures at all? How do you feel about that?
Well, it is useful to--we have the--the evidence is that there's a mixed stain and the profile in the mixture matches that of known people, and it may be sufficient to stop there in saying that there's a match for failure to exclude those people. But it is useful to say if this is an astonishing event or not and it gives us some sense of whether we should be surprised to see such a mixture.
Now, you've described the fact that you had previously not testified in any other case to these approximate frequencies for mixtures; is that right?
In this particular case, were you asked to calculate frequencies for the mixtures in these results?
Were you in fact--well, let me rephrase that. Did you in fact calculate these frequencies for these mixtures--and I'm talking about now the stains that have more than one source so that we're clear about which samples we're talking about.
Were you in fact asked to calculate mixtures for those frequencies as a result of a direct order of the Court?
Have you been previously asked by any courts or ordered by any courts to present frequencies for mixtures in any other case you've been involved in?
And, your Honor, with the Court's permission, I would like to use the glove results board for a moment. For the record, your Honor, that's exhibit 272-B.
And in particular, if I can draw your attention, Dr. Weir--and I don't know if you can see it--but there is at the very top--
In particular, at the very top is an item described on the chart as "Inside back of wrist and back below little finger." Do you see that?
To your knowledge, is that the only mixture that you were asked to calculate a frequency for that came about as a result of testing by the Los Angeles Police Department?
Yes. And that's indirect knowledge. I think it was you who brought that to my attention.
Yes. And when were you actually--well, were you--I'm sorry. Let me rephrase that. Were you asked by me to calculate a frequency as to that particular sample?
That's right. I'm not sure if it was Wednesday or Thursday. It was one of those two nights.
Is that--well, let me rephrase that. Is that the only sample that you were asked to calculate a mixture frequency for from the LAPD?
With the Court's permission, I would like to put up the Bronco automobile result board.
And referring you--first of all, Dr. Weir, do you have a smaller version of the Bronco results board also?
In particular, I'd like to refer your attention to item no. 31 labeled the center console. Do you see that?
You have described previously that you have reported frequencies for that mixture based on the 1.2 allele being actually present; is that right?
You described both I believe last week as well as earlier today that you, when you calculate your frequencies, calculate those numbers based on the types that the analyst reports as being present.
In that particular sample--and let's start with the results board--does that result in fact indicate the 1.2 is present?
Yes, it does. It has--does not have the word "Possible" or any other descriptive there. It's quite clearly there just by itself.
Did you take any steps to verify the presence of that 1.2 allele by any conversation or communication with any testing analyst?
Well, I noticed that it was different from other instances where the 1.2 was written with "Possible." So I checked with Gary Sims.
As a result of that conversation, did you then--well, let me rephrase that a little differently. As a result of that conversation, did it verify in your mind the presence of the 1.2 allele in that sample?
What I'm actually referring you to, if I may, Dr. Weir, as a result of that conversation, did you then calculate--or actually let me rephrase that. Did it verify in your mind that your previous calculation was appropriate?
Is it then correct that with regard to item no. 31 and the frequencies that are written off to the far right, which are for both two contributors, three contributors and four contributors, are then accurate?
Well, they haven't changed since Friday. So I have in my--well, maybe I should look and see what's there.
These seem to be consistent with what we said on Friday. These results haven't changed.
In other words, there was no need in your opinion to change any of the numbers on item no. 31 based on your conversation with Gary Sims as well as the data that had been presented to you before?
Now, I'd like to turn your attention, Dr. Weir, if I could to item no. 29. Do you recall Mr. Neufeld asking you some questions towards the end or very end of his cross-examination?
And in particular, do you recall questions he asked you about whether or not the 1.2 allele in the sample no. 29--and I'm referring to the DQ-Alpha results--may or may not have been present? Do you recall that question?
Now, with regard to that sample--and let's refer to the chart itself--there are actually results on item no. 29 at the DQ-Alpha marker from two different laboratories; is that right?
Let's focus initially on the Department of Justice's results. That's the upper most of the two laboratories that are listed to the right of item 29?
Those results, do they make any statement about whether or not the 1.2 allele is present?
Referring you to Cellmark's results as well on that same item, is there any qualifier about the presence of the 1.2 allele for item no. 29?
I'd like you to assume hypothetically that the analyst testified in this case that with regard to item 29, the 1.2 allele is present based on their interpretation of the data. Would your frequencies under that hypothetical that you have listed off to the right--
Have you made a recalculation of a portion of item no. 29 as far as the frequencies themselves?
That's a different issue. It's--are we allowed to multiply together all the numbers in the profile, and there's a hint of some dependence in one of the FBI's databases, their black database. So for any calculation involving black database, I--oh, I'm answering too much. That refers to the polymarker. Excuse me. The DQ-Alpha I did not change. I'm sorry.
In your view, as a result--and again, assuming hypothetically that both analysts testified to the presence of the 1.2 allele, is there any reason that you should assume it is not present?
Now, your Honor, what I'm going to ask the witness to do is, with respect to certain calculations, we have prepared numbers in accordance with the recalculations already described by the witness, and I have magnetic boards that I would like to use with numbers already written on them.
All right. Mr. Clarke, why don't you start with the ones Mr. Neufeld already has had a chance to look at.
First of all, with a number of these mixtures that you described earlier today, have you had an opportunity to recalculate them over the weekend?
First of all, with regard to--and let's start initially with items 303, 304 and 305 that are from the center console of the Bronco.
Those all actually in your initial or original testimony about frequencies under these various possible combinations, those numbers were the same for all three of those items; is that right?
As far as the calculations you described in your testimony last week, did you commit an error in making those calculations?
Is it the case that what you left out was the assumption that the 1.2 allele for these three samples might not be there?
As it turned out, that was--I thought I was doing it correctly, but the way I had written--my computer program did not fill that in properly. I thought it had and it hadn't.
The effect of assuming that the 1.2 allele may not be there, how did that impact your calculations? If you can describe that in a simple fashion.
No, sir. I left it out for all the profiles that involved the 1.3 and the 4. Every time I should have made this accommodation, I failed to.
It was so obvious, there was no way. In fact, I--I've pointed out now that I did additional items.
KEY QUOTEIf outside of court Mr. Neufeld had pointed it out to you, that it was missing, would you have corrected it then?
Would you have included it if you yourself had discovered it missing in reading your report?
Have you since last week made that calculation or those calculations for these three samples, 303, 304 and 305?
Have you also made those calculations for a few other samples that the same occurrence happened in?
Now, referring you, if I can, to what appear to be three magnetic covers labeled 303, 304 and 305. Those appear to have numbers written in the same manner as what's currently on the Bronco results chart for the possibility of two contributors, three contributors and four contributors; is that right?
And do those three magnetic covers for items 303 through 305, do they appear to have the corrected numbers accounting for the possibility that the 1.2 allele was not present?
With the Court's permission then, I'd like to apply those to the board bearing in mind they can be removed because they're magnetic.
Now, Dr. Weir, what I'm going to ask you to do is, as far as both the old and the new calculations for--and let's just take 303. Do you have each of those numbers in front of you so that you don't have to try to read the board from an angle?
Now, the original numbers reported--and I'll use 303--for the assumption of two contributors in the mixture was 1 in 1,400 as the most common; is that right?
Well, they're different numbers obviously. One's about little over twice the size of the other one. They both convey the impression that the profile is not especially rare. It's--I don't know if you have to have a particular good--I don't have a particularly good word to describe it, but it has the frequency in either the hundreds or the low thousands, whichever way we go.
Okay. The least common for two contributors is 1 in 18,000; is that right? And I'm referring to the frequencies you provided in Court last week.
And is the new figure at the upper range, that is the least common, approximately 1 in 10,000?
Would your comments about the difference between 18,000 and 10,000 be consistent with what you just described about the difference at the most common number?
And when you comment about 303, your comments, do they not, apply equally to 304 and 305 as well?
Now, first of all, in terms of the actual results of the DNA tests themselves, there are three individuals in this case who are not excluded from the actual tests themselves on 303, 304 and 305; is that right?
The original frequencies that you reported for the assumption of three contributors last week, the more common number or most common was about 1 in 450; is that right?
You have recalculated, and that new number, instead of 1 in 450, is approximately what for three contributors?
Objection, your Honor. I would ask that if he do it mathematically, that I would not allow him to give opinions because the frequencies speak for themselves.
Well, the numbers are different. They differ by 150. That's less than a factor of two. I don't think that--I think both the numbers are conveying the same sense of the frequency of this mixture of three genotypes.
The original number--and we'll look now at the upper end, that is the least common frequency--
--for three contributors again, appears to read 1 in 29,000 from last week; is that right?
It's difficult to say that they would give us a different impression. They are both saying the same thing. This mixture would occur with the frequency in the thousands.
And then lastly, on the assumption of four contributors, the range--and perhaps we can do this a little quicker--the range that you provided us last week for the assumption again of four persons contributing to that mixture was between approximately 1 in 49 to 1 in 69,000; is that right?
And the new figures provided today are between 1 in 41 to 1 in 65,000; is that right?
Would your comments about the differences between your original numbers and the numbers today that you described for two and three contributors be consistent with four also?
Now, while we're on the results from the Bronco, if we could, have you also prepared a magnetic cover reflecting what should be or what are the approximate frequencies for two, three and four contributors for the steering wheel, item no. 29?
And showing you a new magnetic cover, would you take a moment to look at those and tell us if those are in fact accurate estimates?
And as far as the original number for two contributors, that was from 1 in 60 to 1 in 11,000; is that right?
As far as assuming three contributors, the original estimate was 1 in 9 to 1 in 3500?
And then lastly, the old or former estimate was 1 in 1 to 1 in 3,000, and now it is 1 in 1 to 1 in 1300 approximately; is that right?
As far as these estimates, Dr. Weir, that you provided last week and in some cases provided additional numbers for those samples this week, do those apply only to the mixed samples in this case?
So in other words, your comments today about different numbers from the same item number have nothing to do with the estimates calculated by the laboratories for stains from one source?
And referring you, if I can, Dr. Weir, to that portion of the glove labeled area G10, do you recall that?
As far as your calculations last week, was this another example where you did not consider the possibility that the 1.2 allele on that sample may not be present?
Did you perform a recalculation similarly to that that you did with items 303, 304, 305?
--and what appear to be frequencies or estimates for either two contributors or three contributors, do those estimates reflect the calculations you've performed over the weekend?
Now, the original numbers on 310--I'm sorry--area G10 on People's exhibit 272-B were from 1 in 3900 to 1 in 19,000?
Assuming three contributors, the most common number went from 1 in 4600 last week to 1 in 3200 today?
And as to the rarest frequency, initially, it was 1--approximately 1 in 71,000, which is now approximately 1 in 63,000?
Well, they're real. They're obviously different numbers. I don't think how--I don't think we get a different impression of their frequencies. They both have--both tell us the frequencies are in the hundreds to thousands.
Now, turning, if I can, to the glove. And I'm referring back to the first sample, that tested by the Los Angeles Police Department. In other words, what's listed as the first actual item on this chart. Do you understand which one I'm referring to?
And showing you what's listed as another cover--rather on which is listed frequencies for two or three contributors, do those appear to be accurate?
Perhaps you can tell us--if I just ask you to describe, can you tell us the difference between those numbers and what is currently listed on the board from your testimony last week?
Why don't I just read from the board? 1 in 30 to 1 in 190. Do I give the new figure?
So the 1 in 30 becomes a 14 and the 1 in 90 becomes a 60. The three contributors, the 1 in 6 becomes 1 in 4. The 1 in 90 becomes 1 in 180.
And now if we can deal with one in a set of three again, with regard to G1, G2 and G4, those three samples contain the same set of types as far as the PCR results; is that right?
Okay. Showing you covers for G1, G2 and G4, could you take a moment to look at those three.
These recalculations were made as a result of your discovery of what that impacted the frequencies or the estimates you provided last week?
This is the same phenomena. This is the failure to account properly for the DQ-Alpha type.
Can you describe for us--and let's turn to the original estimates for these mixtures on G1 or this mixture. Your original estimate for two contributors was approximately 1 in 600 to 1 in 11,000; is that right?
And is it correct that for that same assumption of two contributors, the estimate is 1 in 240 to 1 in 6300?
As far as three contributors, your initial estimate was approximately 1 in 400 to 1 in 36,000?
As far as the differences between these numbers and those described by you last week, would your comments that you made earlier be the same or apply here as well?
And, your Honor, with the Court's permission, I'm going to put the cover on each of G1, G2 and G4, but only the PCR portion.
Dr. Weir, incidentally, as far as these magnetic covers, they can be removed to see what's underneath them; is that right?
Okay. And in fact, it does appear that I'm able to remove them and put them back. Turning your attention now to G3 and G9, do those two samples also go together as far as calculating these approximate frequencies for PCR mixtures as to those samples?
DR. WEIR: Yes, they do. (Discussion held off the record between the Deputy District Attorneys.)
Your initial calculation on G3 was approximately 1 in 3900 to 1 in 22,000 for two contributors?
And the frequencies that you have provided today, instead of 1 in 3900, the more common would be or most common, approximately 1 in 1600?
And as far as the initial calculation for the high end, assuming two contributors--
Turning to three contributors, that range as far as G3 again was originally from 9,000 to 150,000?
And lastly, as to the glove, Dr. Weir, showing you what's labeled G9, these are supposed to be the same numbers as G3; is that right?
Now, using this example, before I put the cover on, Dr. Weir, it appears, for instance, that with regard to the assumption of three contributors, the calculations that you've made and described just now in court, that the upper range is even rarer than the original range that you provided for three contributors on this sample; is that right?
Yes. It's just more difficult to arrive at that--that range. It's a little difficult to make general statements about how the numbers will come out, but it's a rare event when I do it properly.
So, in other words, it's not necessarily the case at all that when you make your calculations based on the assumption the 1.2 allele may not be present, that in fact the range may show us that it's something that's even a rarer event?
Is it the case, Dr. Weir, that when you perform your recalculations, at least one of the ends of the range may be rarer than you calculated it last week?
And lastly, your Honor, I'd like to refer to--as far as these mixtures, to the Bundy results board.
Dr. Weir, referring you to item no. 78, and in particular, the PCR portion of the frequencies off to the right of that particular item?
Did you also recalculate the possible range or estimates for the range of the characteristics from that particular item based on the PCR markers only?
And have you calculated those and are they accurately reflected on another magnetic cover labeled item no. 78, PCR?
Could you tell us the original range and that reflected on the magnetic cover as well?
Let me read from the board. So 78, the--we had the hundred--the 1--looks like--your writing is worse than mine, Mr. Clarke.
Looks like 1 in 285 to 120,000. That becomes 1 in 110 to 1 in 16,000. So three contributors, 1 in 60 to 1 in 490,000 becomes 1 in 42 to 1 in 44,000.
Could you verify in your calculations--can you tell us whether the old number for two contributors was 1 in 285?
You'd just have to give me a minute. That sounds right. I need to check. Item 78 was--yes, the lower end was 1 in 285. 285.
And as far as the three contributors, what would be the most common? In other words, if you could tell us the old and the new number as far as three contributors in the most common end of the range.
Dr. Weir, could you describe for us approximately how many calculations you've had to do in this case? And I'm referring to--and I'm having you, if you would, set aside recalculations. Just the calculations you did initially?
Everything. I don't think so. There are many, many items of evidence on these charts. Many of them have multiple loci, and I've done a lot of things with them. So certainly several hundred calculations.
As far as these frequencies that you have recalculated, those involve mixtures alone; is that right?
In terms of these frequencies on the mixtures and on the current board, the Bundy board, item no. 78 appears to be the only mixture; is that right?
As far as those frequencies are concerned, do they take into account in any manner any of the known types of either Mr. Simpson, Nicole Brown or Ronald Goldman?
No, they don't. None of the calculations I do take into account the included people.
KEY QUOTEWell, we have--we have an item of evidence and a DNA profile which matches some known people. And we set that aside. We say there's a match. Well, I mean, the investigation proceeds because there is a match. Having the match--once the match has been declared, that is set aside for all these calculations, and then we say where we have this profile, how likely is that profile if there were unknown contributors, people we don't know anything about. So all these numbers refer to the unknown people.
So, in other words, your frequencies for all of these mixtures are calculated without regard to whoever they may or may not be consistent with in this case?
As far as your recalculation--well, I'm sorry. Let me rephrase that. These calculations, not taking into account any of these known people, is that different in any way from the calculations done for a single source stain; for instance, anything but 78 on this Bundy results board?
No. The single stains are just the same. If there's a single stain, it has a profile which matches a single person. That person is set aside, the calculation is done to get the chance of getting that profile typed if there was an unknown contributor.
And again, going back to the mixtures then, for example, on just the PCR results alone on item no. 78, when you say that if there are two contributors, the frequency is approximately 110 to 16,000, that's just looking at two contributors at random on the entire planet?
I think they're totally misleading. They completely ignore the fact that there is a mixture.
That's what makes me so angry with myself.
It was so obvious, there was no way. In fact, I've pointed out now that I did additional items.
your writing is worse than mine, Mr. Clarke.
None of the calculations I do take into account the included people.