All right. Thank you, ladies and gentlemen. Please be seated. Dr. Gerdes. All right. The record should reflect we have been rejoined by all the members of our jury panel. Dr. Gerdes is again on the witness stand undergoing cross-examination by Mr. Clarke. Mr. Clarke, you may continue.
Thank you, your Honor. Good afternoon again, ladies and gentlemen.
THE JURY: Good afternoon.
Dr. Gerdes, I would like to shift your attention to cross-hybridization, and in particular, with regard to cross-hybridization that can be characterized, can it not, as the appearance of additional alleles because of the closeness in sequence of one to the other? Is that a fair definition?
And is it correct that in that report you have described the fact that additional alleles can appear weaker than alleles from the sample itself when, for instance, the temperature of this process of the typing strips showing these dots isn't correct?
That's correct. If either the temperature or the salt concentration is incorrectly--the protocol is incorrectly followed, that would be an explanation--
The temperature could be off during this process and you might see a weak dot; is that right?
Or how much salt is in the solution that this development process goes on in, that could also result in a weaker dot?
It might--I have to clarify that. It would result in a weaker dot on a sample from a known individual. It would not--there is no way it could result in a weak dot on a no DNA control such as the amplification blank or the reagent blank or a substrate control where there is not supposed to be any DNA in the first place. It could not result in a dot there.
In other words, when we put that chart up there of controls, if that were a reagent blank, also called a cloth control and you saw a weak signal on one of those dots, that is not from cross-hybridization, correct?
But on, for instance, a known sample from a person like the positive control that we described there, that could be the subject to these additional dots showing if, for instance, the temperature were off or the salt solution were off?
Now, as far as the differences between these alleles, and let's use DQ-Alpha, there is a 1.1 allele; is that right?
As far as the 1.1, allele it is very close in terms of its DNA sequence to the 1.2 allele; is that right?
And in fact there is only one little piece of DNA information, a base pair difference between the 1.1 and the 1.2, correct?
It is that similarity in closeness that can lead to the appearance of these additional weaker dots if a sample is, say, a 1.3 and you see a 1.1 showing up weekly?
Well, specifically in the case of the 1.1, that is not really the reason that has been described for that particular allele.
Okay. We will return to that. But the 1.1 and the 1.2 only differ by one base pair, correct?
And a 1.2 type can also show--simply because of this one base pair difference show some weaker reaction in the 1.1, correct?
If the reaction and the protocol are correctly followed, this system is capable of detecting a one base pair difference.
It is only when you have larger amounts of DNA that you have a minor percentage of that kind of cross because of the one base pair mixing.
Isn't it true that this cross-hybridization due to this one base pair difference between a 1.1 and a 1.2 can appear even if large amounts of DNA are not involved?
It is--the literature that I have read suggests that six nanograms and above is where the problem is.
Is it your testimony that unless there is six nanograms this cross-hybridization we know as 1.1 and a 1.2 can't happen?
It is possible, but it is not the more likely explanation when the DNA concentration is low.
Now, the 1.1 and the 1.3 alleles also differ by only one piece of DNA or base pair, correct?
In other words, these 1.1's, 1.2's and 1.3's are very close to one another in their sequence?
Again this system is designed to detect that minor difference, one base pair in this kind of system is significant.
As far as this one base pair difference, isn't that an important piece of information for an analyst to know?
Now, in addition to your report, the user guide talks about this very cross-hybridization, doesn't it?
They also describe--well, let me rephrase. The package insert that also comes with the kit also discusses this minor difference in sequence between these three subtypes of the one, correct?
Incidentally, do you encounter any cross-hybridization in your own laboratory work?
Whenever you use a dot-blot format this is a fact of life. You are going to see a certain degree of cross-hybridization.
And is that true in your laboratory's casework? I'm referring to samples that come into your laboratory you perform testing and you report results to doctors to talk about with their patients?
Yes. We deal with it in a little different specific manner to accommodate that problem.
Is it also something that in your casework as an analyst performing this type of work you have to be aware of?
And is it something that from that awareness and experience you use that awareness and experience to properly interpret results?
Yes. What you have to remember is in our particular case we are dealing with known single individuals. The problem becomes acute or exacerbated--
Now, let's turn to this DX. In fact, what you've referred to as DX is the fact that a 1.1 allele may appear weakly, that is, weaker than the other alleles from a sample in a given case because that--because of the--I'm sorry, let me rephrase that. This DX gene phenomenon you have described is due to the fact that there is a gene close to DQ-Alpha, right?
That's correct, close, not--not only in proximity, but close in terms of its sequence.
And some individuals who while they don't have the 1.1 DQ-Alpha type, may have a form of the DX gene that will weakly light up that 1.1 dot, correct?
You have recognized this DX gene, 1.1 reaction in this case by way of your report, correct?
Now, you have described the fact that in your opinion a 1.1 reaction can only be as a result of this DX gene if the person doesn't have a 1 type at all. Do you recall that?
Okay. Isn't it correct that there can be DX gene reaction from a person who is a type 1.1, 1.2 or 1.3?
That is--that is true, but you can no longer differentiate as to whether it is true DX or if it is a true 1 type because that 1 dot is lighting up.
Okay. In a sample where the 1 allele is lighting up and let's say the person is a 1.3 and a 3, let's say, the 1 dot lights up, right?
It can be, but you can't tell that that in fact if that is from a contaminant or if that is from DX because in that situation you don't--you have the 1 dot lighting up, so you have no way of knowing one way or another.
What you are saying then, Dr. Gerdes, is the fact that if you see the 1.1 lighting up weakly, but the 1 dot isn't lighting up at all, you would have more confidence in concluding that that is from DX gene activity?
No. I'm saying with any degree of scientific certainty you can't say one way or the other.
But in my scenario wouldn't you have more confidence that a 1.1 reaction is due to DX if you see no reaction whatsoever in the 1?
And with regard to that, isn't it correct that with regard to mixed samples the more genetic markers you look at the more information you can obtain about a mixture?
In other words, if you are able to look at seven genetic markers and across all of those markers the results are consistent with a mixture of two people--
--two known people, that those seven markers give you more confidence in who that mixture came from than if you did simply a DQ-Alpha test alone?
As far as mixed samples--well, let me rephrase that. As far as contamination is concerned, isn't it true that if you use a second gene, a third gene and a fourth gene, that the likelihood of all of them being contaminated is pretty low?
Again, it depends on the source of contamination. If the contamination is due to cross-contamination, that is the transfer that we talked about early on when samples are handled, that is going to be repeatedly typed, no matter how many genes you look at, as the same DNA.
In other words, the more markers you look at, the more you can eliminate carry-over contamination having any impact at all or even occurring?
It--that is true, but it depends on the level of background contamination in the lab at the time in terms of how much information you can get out of that approach.
Every marker you add, when the results demonstrate consistency from, say, a single person or two people, gives you more confidence that carry-over contamination didn't happen?
On that specific sample, but the--the point is, if you have evidence of this kind of contamination at a high level background level in the laboratory, then that analysis falls away. I mean, it is harder to do it under those circumstances. It is sort of like if you went into your house and you noticed a cockroach and then the next day you noticed another one, it would be an indication to you that maybe there are other roaches around and you maybe better do something about it.
But as far as this carry-over contamination, each time do you another marker and you see consistency of results from, say, one or two persons, you are gaining more confidence as a scientist that carry-over contamination didn't happen?
Now, with regard to mixtures, isn't it scientifically a good idea to do additional genetic markers to get more information about that mixture?
Repeat testing also plays an important role in interpreting mixtures; is that right?
Yes. Again assuming--it depends upon the stage at which the contamination occurred. If it occurs early on and is carried through a series, repeat testing isn't going to help you, but in the sense of the random events of contamination it would be helpful.
I would like to shift your attention, if I can, Dr. Gerdes, to your review of the LAPD validation strips.
You have described yesterday that from your view the Los Angeles Police Department DNA laboratory has by far the worse contamination you have seen in a forensic lab. Have I accurately described what you testified to yesterday?
Now, the time period that you examined the Los Angeles Police Department's DNA typing results was approximately May of `93 through August of `94?
How many other laboratories have you examined all of their PCR DQ-Alpha typing strips for a 15-month time period?
KEY QUOTEThe sample that you looked at in this case, as far as the individual strips, approximately how many was that?
Yes, counting every strip that you look at, and a rough estimate is fine if that would make it a little faster at this point?
As far as your review of the materials in this case, did you speak to Dennis Fung about his evidence collection?
Did you speak to any of the employees at the LAPD laboratory about their role in this case in collecting evidence?
As far as your review of the strips and, we will return to the numbers in a moment, did you speak to any of the analysts asking them to help you interpret an individual strip in your review?
I--Collin Yamauchi and myself sat together and I pointed out to him the things that I was observing.
Did you ask him questions about how he performed his testing as far as interpreting alleles on strips?
No. I told him my interpretation and he with us present to respond and at that time he didn't.
I told him my interpretation. We were sitting there together and I would say "This is a weak dot that I'm concerned about," and we go through page by page, "This is what I'm concerned about," and there was--I think he made note of those things, but I don't even know if he made official note of those, but he was certainly there at the time.
As far as any scientific test, let's use DNA typing, is it in your view important to learn how to use a test?
Is it important to have actual physical hands-on use, hands-on work using a technique to be able to perform that technique?
If you are going to perform the technique, yes, but in interpretation of scientific data, that is the nature of science. It should result in a documented piece of paper or evidence or result that can be independently analyzed by another individual who is familiar with the way--familiar with the interpretation of those results.
In your view is it important to have experience in actually using a technique to be able to interpret it, as you have done in this case?
KEY QUOTEIn your view then it is not important at all for you to ever have performed one of these forensic analyses?
The--the interpretation of the strips are fairly straightforward. I don't feel that is necessary.
Now, as far as the LAPD's casework, when did it actually begin, in other words, accepting cases, performing DNA typing, that is PCR DQ-Alpha typing in actual cases?
You, as a result of examining these strips that you described yesterday, created a chart; is that right?
By chart I'm referring to a document of several pages cetera in which you note a particular strip on a particular run and then a few comments about that run, correct?
All of them they gave me. I don't think I have the entire database, but in that time frame.
Well, as far as their strips, how are they kept at their laboratory? In other words, what did you physically look at?
I looked at a book very similar to what I have here, which is a run book which shows each--you know, day by day what the run and the strip looked like.
Umm, the page has a photo of the strips and it also has the recorded typing result and the name of the analyst, the date.
Well, I don't know that we need all the details, but is it correct to say for now that there is identifying information about what hybridization record it is?
Yes, there is a hybridization number. There are two numbers and item number, a description of the item and then the hybridization volume, the results and also there are items about the lot numbers of specific reagents that were used to produce that result.
And I believe you described yesterday that you reviewed during this time period of May, `93, through August of `94, all of the LAPD DNA laboratory casework strips, correct?
I--I'm not--I don't believe I was shown all of the casework strips, but all of them that I was shown I reviewed, yes.
There are a few, but I believe they have to do with the fact that they were polymarker runs that were incorporated in here which were given a sequence number and that wouldn't be counted as DQ-Alpha.
Okay. So from this 1 through 261 you would look at each of the pages that involved DQ-Alpha typing and if a page involved polymarker typing you didn't look at it, you would go to the next hybridization as well as--
I looked at some of these as well. I didn't analyze those in as systematic a way as I did this.
So you had an opportunity to look at all the polymarker results in the LAPD laboratory as well; is that right?
Did you in fact create any chart or document about polymarker results by the LAPD laboratory during this August through--May `93, through August of `94, time period?
Now, as far as the term "Run," "Hybridization run," that term was used yesterday by you, correct?
Well, the strip--each strip is one sample and a run would be all of the particular strips that were run on a given day.
So a run may include one page of strips or several pages of strips; is that right?
As far as a run being in one day, what if two separate runs in the thermalcycler were done on the same day, how would you deal with that?
If it was by a different analyst, I would consider it as a different run, otherwise it would be considered as the same.
Okay. So if there is two analysts, one using the thermalcycler and then another one later using the thermalcycler, that is two runs?
I went pretty much by date. One day, even if there were two people, that would still be considered one run, one day; one run.
Okay. So when, for instance, there were charts--and I believe you testified yesterday--you did testify yesterday about contamination by run. Do you recall that?
Yes. It is by date, so everything that was done on a particular day is considered a run.
On a quick look through here I don't see a single day where both analysts ran on the same day.
Okay. At this point it is your testimony from your review of your material briefly that you don't believe there was more than one run by two different analysts on the same day?
Okay. If in fact there were two runs by an analyst, that is by two analysts, two different analysts on the same day, wouldn't your charts about contamination by runs be wrong?
The point really is you are getting confused in terms of--you are making too big a deal out of this "Run" definition. The point behind this is to look at time, what happened on a given day, what is in that lab on that given day, what is in the lab the next day, what is in the lab across time over the course of a month. And so that is the thing that is really critical here in terms of what kind of contamination is occurring on different days.
You've already described on those charts contamination by strips. Do you recall that?
Well, I believe I told--I titled it contamination and artifact by strip and that tells you how often--that gives you an idea or a sense of how often you see additional dots indicating additional human DNA in terms of just looking at every every step and asking that simple question. Is there a dot there that shouldn't be there?
Dr. Gerdes, if your definition is a run is a run done by a single analyst and two different analysts perform runs on the same day and you treat that as one run, doesn't that make your chart wrong?
It may be wrong by one item in the--in that particular entry. I don't believe so. I just checked through here and I don't think there this is a date where a single--you know, two analysts ran on the same day.
No. It is my view that I went through here and I don't believe that there is an error and that a run has more to do--you are confusing it. It has more to do with what is going on at a particular time.
Now, as far as--and this chart you created, the several-page document that you described earlier, you originally constructed that document and named it, correct?
The original name of that document was what, the chart I'm referring to that you've constructed that is several eight-and-a-half-by-eleven pages?
I believe it was the observation of--something to the effect of LAPD DQ-Alpha contamination. I believe that is--something to that effect. I don't have the original chart with me.
All right. Would it be consistent--perhaps if I could refresh your recollection. Would it refresh your recollection to see a copy of that original chart?
Showing you my copy, Dr. Gerdes, does that refresh your recollection as to your original name for your chart?
First of all, the first chart, to your knowledge was that provided to the People of this case?
It was never finalized until just recently when I changed the title. It was constructed over a period of many months.
I--I don't remember exactly when I mailed my first copy of that, but I believe it was around January.
Yes, but at that point it was--it was a working document and it hasn't been--it wasn't finalized until just recently.
Well, as you can see there are a thousand strips I was looking through. I wanted to double-check things. I went through it a number of times. I changed the title. I changed entries. These are things I'm working on.
You looked at that closely to make sure that that document was accurate, didn't you?
You did everything you could to make sure you made no mistakes in that document; isn't that correct?
In fact if that was a working document in January of 1995, if that document was not provided to the People until a later time, you wanted to make sure that that document was a hundred percent accurate?
Umm, it was in early June, I believe. I'm not sure what the date was exactly. Well, the report--I don't have a copy of the letter, but it was not that long ago, within a month.
Okay. Let's go back for a moment if we can. You say this is a working document and you were making changes to it. What changes were you making from January and over the months since January?
The way I constructed that chart was to go through and make the entries and then I would double-check them and I would also--I would--as I went through the first run through I would have questions, you know, have--have them see if they can get another copy of this strip or this one is missing, for instance, the initial discovery that we received was incomplete. We had to have--we had to go back to the laboratory actually in January and get a second set, a totally different set of photographs for strips that were missing. So once I had strips coming in that were missing, I would go ahead and revise the table with regards to strips that were provided later, that sort of thing.
Is it your testimony that all you did after January was add notations with strips you hadn't seen yet?
All right. We will return to those, but as far as this working document and other than the strips that you were given that you didn't have and the changes we will discuss, did you make any other changes to that document between January of this year and June 1st of this year?
Again, when I go through the first time I would make notations, look at this strip. It was a working document. I would go back and as I--I have convinced myself on certain strips, I would either decide not to call that one or to call that one or to try and get a better photograph, double-check, make sure that everything is accurate.
He took photographs of the typing at DOJ and he observed that typing being done, I believe. I'm not sure.
You discussed none of the interpretations that you made of the LAPD results with him; is that right?
What did you rename the chart, and if it would refresh your recollection to look at the new chart--
So you made the decision to change the title "Contamination incident" and change those two words of the title to "Unexpected alleles"; is that correct?
KEY QUOTEAnd that is because there may not be contamination in many of the instances that you describe in this chart, correct?
Well, I believe I described it yesterday. It is--the first step is to just ask the simple question is there a dot there that shouldn't be there?
The second step is to interpret those strips in the context of a given day to see if that can be confirmed as contamination.
Dr. Gerdes, you renamed that chart from "Contamination incidents" to "Unexpected alleles" because there are many instances in your chart that may not be contamination, correct?
Specifically on the 1.1 allele where the question would arise, 7.6 percent of those 1.1 alleles could be interpreted as DX. Specifically in terms of the 1.3 allele, which is the other allele where it is known to have cross-hybridization, approximately 25 percent of those might be explained by that cross-hybridization. On all of the other alleles the 4, 2, 3, 1.2, that--those specific alleles have never been described to have those kind of problems, so they stand as real contaminants.
Dr. Gerdes, as you sit here today, is it your testimony that other than that 7.6 percent regarding the DX 1.1 allele that all of the remaining incidents, strip by strip incidents that you note in your chart, are contamination period?
It depends on the allele. You have to look at it allele by allele. We already discussed the fact that there is an ambiguity if you have a 1.1 and a 1 dot. Now you can't decide. Now, I chose to count--since you can't decide if it is a contaminant or not in this analysis, I chose to count that as a contaminant. You can argue in fact, and you have that that 1.--that particular allele, the 1.1, even those might be ambiguous, they might be DX and that--that is true, but other than those, all of the rest are contamination.
In other words, other than the 7.6 percent--and those are the percentages that relate to your interpretations in DX, correct?
Other than those, it is your testimony today that all of the remaining incidents that you identify in your chart are contamination?
Again, the 1.3 allele, you analyze that the same way on the 1.3 allele. If you have a 1 dot as well as the 1.3, you can't tell if that is cross-hybridization. If you have the 1.3 only, that could be considered cross-hybridization, so that is 25 percent where the 1.3 dot only showed up and not the 1 dot, so those also are in question. But with the exception of those two, all of the rest are defined as contamination.
Is it your testimony that with the exception of those, all the rest of the incidents that you have charted are contamination?
You have described in your review the fact that you looked at photographs of these strips at the LAPD in January, was it, 1995?
Did you conduct a review at that laboratory, that is, in person at the laboratory, on any other date of these strips?
You also described the fact that photographs were taken of the photographs of the strips in their book?
All right. Were those photographs of all of the strips or just ones that you had some concern about getting a photograph of?
They were a subset--initially photographs were provided by LAPD and on the second visit in January we photographed strips that were missing or those that I wanted to see and have rephotographed because their photographs was not of good quality.
You then wanted to ensure you had a photograph of each of the 1 through 261 other than the polymarkers that you didn't review?
Did you then obtain a photograph of each of the strips that involve again this DQ-Alpha typing process?
Now, in this--at the time you renamed the chart, to your knowledge was that provided to the People as well, the renamed chart?
To your knowledge was a document also provided to the People indicating your discovery of errors in your chart?
And in particular, as part of your review of your chart, you discovered two errors that you had made, correct?
They are not really errors. They are recalls or reconsiderations in terms of what I would characterize those specific samples.
Well, in one of those instances you labeled a standard as no. 1 and being a mistyping by the LAPD, correct?
I never recorded it as that. If you look at the original table--table, it was recorded as "Standard one question, question mark, typed as `check this.'"
From your original review, and your initial chart provided to the People, you described six mistyping errors by the LAPD, correct?
When you reviewed your material and renamed and corrected your chart, you described five errors by the LAPD; is that right?
The error rate that you originally described in your first chart was incorrect; is that right?
Not--it is still not--it is not incorrect. It basically--on that particular strip there--if you read the result--excuse me--the result, it is a mistyping.
If I can explain it, there is no C dot. There is a recording on the list from the analyst that says "No C dot." I felt I could see a C dot. If there is a C dot there, that is an error because it is a typeable result and it is an error, they recorded the incorrect type. But because the analyst wrote down "No C dot," I decided to change that and say that was not an error. In my opinion it is still an error, but I decided to change that because of the fact that I knew there would be an argument about whether or not the C dot was there and the fact that the analyst recorded there was no C dot; therefore, the analyst probably would not have considered that as a typeable result and would not have reported a typeable result that was an error. But in fact I feel I can see a C dot on that strip, and if there is a C dot there it is an error. If there isn't a C dot there, it is not an error.
So Dr. Gerdes, in your first chart you recorded it as an error? In your renamed altered chart you recorded it is as not an error, but your testimony is it is still an error? Is that fair?
As far as your chart, can you tell us how many actual records there are, hybridization records, the one page that may list eight samples or up to eight samples that you actually listed on your chart, and approximate number is fine?
Okay. You are actually referring to something different, aren't you? I'm asking how many hybridization records, how many different single pages with up to eight samples on the page?
Oh, this? 200--well, somewhere around 200 probably. I don't know exactly how many.
You are asking how many specific pages there are with strips on them? I haven't counted those so I don't know.
You know, they are numbered from 1 to 201 and some of them the hybridization names are recorded on more than one page, so it is hard for me to estimate unless I count these.
Okay. And you only recorded, as far as a record, a hybridization record, where you thought that something was a 1.1, correct, as far as those samples?
Of the samples not recorded by you on your chart, is it your opinion there was no contamination revealed whatsoever?
What you did was record on your chart the occurrence of additional potential contaminant alleles, correct?
These are additional potential contaminant alleles, not contamination for sure, correct?
Incidentally, this typing process included work done by analysts unrelated to this case, correct?
In other words, your survey was of every analyst who conducted typing in the Los Angeles Police Department during that time period?
Dr. Gerdes, did your review include work done by any analyst in the LAPD laboratory, whether connected to this case or not?
Incidentally, the polymarker results, they have controls as well on them, don't they?
The polymarker has an s dot for determining the amount. The difficulty with the polymarker system is most of those are only two allele systems, so you don't have the luxury of looking for more than--I mean, there is only two there, so you can't look for more than two, so the same kind of analysis is more difficult.
With the exception of two genes that are on there that have the three A, B, C genes.
In fact, those controls can be looked at to show if they reveal contamination, can't they?
Wouldn't it be scientifically appropriate to look at the polymarker results as well if you are trying to find out how much contamination is in that laboratory?
So you felt you didn't have sufficient time to look at the entire records of the LAPD DNA typing laboratory?
I felt that the DQ-Alpha--and I looked at spot checks on some of those and there were some of the same kind of things, and I did not have time to do a detailed analysis on the polymarker.
Wouldn't it be scientifically proper or wouldn't it be more appropriate to look at all of those 261 records instead of something over 200 of them?
Well, the polymarker is not--they don't have normally as many strips on those either and it wasn't started until just more recently.
I would have to look at the record to see if they did any polymarkers in that time frame. I don't remember them doing a lot in that time period.
Now, what I would like to address, Dr. Gerdes, in the remaining minutes that we have today--I'm sorry?
The errors that you have described, and let's start, if we could, with the first error, and I'm going to refer to them chronologically if I can.
In chronological order if that is okay. Your first error identified relates to hybridization record no. 15; is that right?
Your Honor, we don't have duplicate sets. May I stand by the witness to look at them?
Your Honor, I have a document that I would like to have marked as the People's next exhibit.
Your Honor, we want to compare the pictures because they are two different generations of photographs.
Does that appear to be a photograph of--not just the photograph of that particular hybridization set of strips, but also of the actual page itself, the document?
And does it appear to be an accurate reproduction and contain the same documentary information as what you have in your own notebook?
As far as the photograph is concerned, does it appear to be approximately the same--the same as the photograph that you have?
Yes. I mean, I have a Xerox here on mine at the moment, so--which is bothering me, but that seems to be the case.
If you don't have a photograph, then what I can simply ask you to do is to look at this original on the elmo, with the Court's permission?
Your Honor, with the Court's permission then what I would ask to do is to place People's 558 on the elmo.
Now, I don't intend to ask Dr. Gerdes about dots and intensities, and perhaps if there is anytime you feel it is important to look at them, please tell me for purposes of this questioning.
You have documented in your analysis one of the samples in this particular run as being one of the five errors that you testified to yesterday, correct?
And, first of all, just a little bit about this record. This is a run that occurred on what date?
So someone who didn't actually perform the actual bench work but rather is asked to come in and look at the strip results?
Now, this particular run on 7/14 of `93 was about eleven months before the incident in question in this case?
And would this then be during the fairly early stages of the LAPD's use of DNA typing?
Now, in particular, the error you identified--and I'm going to ask if we can zoom in on the first row not of the photograph but of the actual written documentation. Actually, if we can go up and zoom out a little bit. Perhaps just a little bit more. Perfect.
The error that you identified is actually on the first sample listed on the chart which is labeled on the left no. 1, correct?
Which has identifying information and then a description of the medical sign for female reference blood?
And the results of that test by Erin Riley she has written down on the far right "1.3, 4," correct?
Now, in the course of that run there was activity in one of the negative controls; is that right?
And in fact written down, and perhaps we can zoom in a little bit on the very lower right-hand corner of the eight-segment graph--
That's correct, on the cloth control, the substrate control, the extraction control.
That is a control you described earlier that in the course of typing that is supposed to be negative, correct?
As far as that control is concerned, if that is not negative, the result shouldn't be reported, correct?
The only showing of those alleles that is the call on that reference blood, is the 1.3, 4 that is written under dq-A1 results, correct?
I'm not sure if that is the reason she did it. I don't know why she did it, but she did rerun the sample.
Wouldn't it be scientifically appropriate to rerun that sample because of the control that showed activity?
It would be scientifically acceptable to begin with something that hadn't even been exposed to that laboratory and rerun it.
Dr. Gerdes, that 1.3, 4 that is written in by Erin Riley, that is not the correct type for that sample, right?
When Erin Riley reran that sample, if she reran that sample, that was the proper thing to do, wasn't it?
All right. Your Honor, I have a second similar photograph I would ask be marked People's 559.
Dr. Gerdes, showing you what is labeled at the top "Hybridization 19," is that a similar photograph of a hybridization record and photograph attached to that record as was the case with the previous exhibit that we've discussed?
And does it appear--contain the same information that you have in your own records?
All right. With the Court's permission I would like to also place that on the elmo.
Actually just to be clear, Dr. Gerdes, that previous photo, exhibit 558, Erin Riley called the dots, the probes, the alleles, that she observed in that typing strip, correct?
And as far as her noting the 1.3 allele and the 4 allele, your review also revealed that the 1.3 allele and the 4 allele were present, correct?
Now, looking at the next exhibit, 559, and if we can zoom in on the same area, let's start there. This is a similar record, correct?
And in fact on this record--your Honor, I have skipped one exhibit. I'm going to have to ask that another one be marked as what would be People's 560.
Yes. Dr. Gerdes, showing you, if I can, what will be People's exhibit 560, does that appear to be a similar photograph of a record and photo?
Now, as far as this particular record, which is People's 560, does that appear to be a record from the next day, July 15th, `93?
Now, on this particular set of hybridization records what location was that sample that was rerun from the previous day?
One that was called an error by you? In other words, is that sample rerun on this exhibit?
And on that occasion she noted the alleles 1.3 and 4 being present; is that right?
And that notation is the first notation in the immediate--I'm sorry, the far right column that is labeled "Dq-A1 results," correct?
And in fact what is noted for the activity seen by both Erin Riley and Collin Yamauchi with that negative control?
In view of the existence of that activity in the negative control, isn't it scientifically proper to rerun that sample again?
It is an error in the sense that that contamination in the laboratory created a false typing result.
Is it your testimony that from this document Erin Riley conclusively established, for purposes of reporting results, that that was a 1.3, 4?
No. My testimony is that that result was recorded as a type that was an incorrect type as a result of the contamination that was to found in the lab.
Wasn't Erin Riley--let me rephrase. Didn't she rerun this sample again because of that activity in the negative control?
And in fact did she rerun that sample as reflected in what is People's exhibit 559 which I showed to you a few moments ago?
Incidentally, before we do that, Dr. Gerdes, you are not faulting Erin Riley for noting the alleles that she saw in that sample, are you?
All right. With the Court's permission I would like to again place on the elmo People's 559.
Now, lastly, Dr. Gerdes, with respect to this series of runs, this was a run conducted by Erin Riley on or two days later, July 17, `93, correct?
Now, on that chart, if we can focus in just on the graph portion of the chart, can you tell us, doctor, where that sample is on this chart?
And then reading across to the right--this would be three rows from the bottom, correct?
Well, that stands--that is FID is an abbreviation that is used in the proficiency testing, so this is actually a proficiency test that they were sent a sample and that is just one of the labels for that sample.
All right. And over to the far right of that same sample and this has been run for the third time by Erin Riley, correct?
Looking at the cloth control, the negative control, what activity did that negative control show?
Based on all of this information, Dr. Gerdes, is it your opinion that Erin Riley and the LAPD laboratory made two errors?
Dr. Gerdes, is it your testimony that based on the first two runs before this one that those were two errors by Erin Riley?
Incidentally, if such an incident occurred in your laboratory with activity showing in the negative controls, would you want your analyst to rerun that sample?
So Erin Riley did just what you would want her to do with these samples or this sample if she were working in your laboratory?
Dr. Gerdes, showing you what is labeled "Hybridization no. 33," does that look familiar?
All right. Your Honor, with the Court's permission I would like to place People's 561 on the elmo.
And that date again precedes the LAPD beginning casework, actually accepting cases in their DNA laboratory?
Now, this particular run--actually this particular record involved the analyst--or involved Collin Yamauchi as the analyst, correct?
Now, in this particular sample, this reflects information from what you--from which you determined that there was a third error as far as mistyping in the DNA laboratory at the LAPD, correct?
Now, the particular sample that you identified as being an error is which one on this particular record?
You are referring to on the actual graph with written information "Sample no. 2," which would be the second row down under "Description"?
Well, wasn't this in fact one of a series of vaginal swabs that were tested using the DQ-Alpha system?
And a series of vaginal swabs that were tested basically as part of the validation process within the LAPD laboratory?
This process whereby analysts tested a number of different types of samples to ensure they were able to use the system properly?
Now, it is labeled--and let's start on the first sample because that will clarify what the second sample is, correct?
Okay. We will talk about that in a moment. Going now down to the sample that you have offered the opinion is an error, that relates to the same swab, correct?
Now, with regard to this particular sample, that is designed to be a vaginal swab for purposes of this test?
And the actual sample that you have offered the opinion is an error, the SC sample, is supposed to be the sperm cells from the male who engaged in intercourse with this female; is that right?
Now, the typing results obtained by Mr. Yamauchi as to the epithelial cell portion are 1.2, 4, in the far upper right-hand corner of the graph?
And in fact that properly shows what the DNA type is of the woman from whom that swab was taken, right?
So that what the EC portion was showing was what was expected to be the portion from the person the swab was taken from, the woman?
And in fact in this typing process that has been previously described in testimony in this case it is a feature of DNA that even though a sample in a sexual assault case may include DNA from the victim and DNA from, for instance, a rapist in a sexual assault crime, that DNA allows to a large extent separation of the female's DNA from the male's DNA; is that correct to say?
By the process of a differential extraction there, without getting too deeply into that, that is a process by which the DNA analyst can break open just the victim's cells and hopefully not break open any of the sperm cells, right?
And then go back and by some other chemical steps break open the sperm cells, set that aside and type that separately, right?
And in fact in a sexual assault case it is not uncommon at all to find, for instance, in the epithelial cell portion some sperm DNA?
Because during this process of breaking open the victim's cells some of the sperm may break open and release the DNA as well?
And isn't it also correct that the opposite occurs, which is in breaking open the sperm cells some of the victim's cells didn't quite release the DNA in the first step and then they are left in the sperm fraction--I'm sorry, the sperm cell portion also?
Now, in this particular instance these results reflect the victim's DNA in both; is that right?
In other words, the 1.2, 4 is consistent with the victim in the epithelial cell portion as well as the male sperm portion?
That event, in the context of a mixed vaginal sexual assault sample, is not unexpected, is it?
And in fact a trained DNA analyst would not be surprised to find exactly what the results from Mr. Yamauchi revealed, correct?
In terms of the epithelial fraction, that's true, but in terms of the sperm fraction that is not true.
Well, a trained analyst would look at those results and first of all, let's see, what is the type of the male donor of the sperm on that sample?
You have reviewed cases involving sexual assault samples as part of your work for criminal defendants, correct?
And you have seen samples in which there are mixtures of victim's DNA in a vaginal swab along with a sexual assault or rapist's DNA, correct?
And in fact you have seen samples in which the victim's DNA shows up in both portions, correct?
A trained analyst, an experienced analyst, even though seeing that 1.2, 4, in some of the cases that you've reviewed, would not exclude a person who is a 1.2, 1.3 from having, I'm sorry, contributed any DNA to that sample, correct?
Is it your view that a person can be excluded as a possible donor when their DNA just doesn't show up at all?
You don't know. That is the problem. The 1.3 has been missed here. If this typing method is reliable, you can't miss things. If you miss things, how do you know that you didn't miss things on other samples? I mean, you can't do that.
Dr. Gerdes, with respect to a rapist, isn't it the case that in some cases they don't leave enough DNA?
That is true, but in the case of a differential extraction the analyst is supposed to look at the specimen and observe sperm before they go ahead with the differential extraction, and if there is sperm there you should find the allele from the sperm and in this case that didn't happen.
Well, is it your testimony that a trained forensic DNA analyst would look at those results and conclude that the 1.2, 4, came from somebody other than the victim?
And they wouldn't conclude that a person who is a 1.2, 1.3 couldn't have raped that person?
If you had--I believe there would be an interpretation problem, but they would most likely not, you are right.
And that is the problem. The person was a 1.2, 1.3, so it is an error that person would not have been--
Dr. Gerdes, a trained analyst would look at those results and they wouldn't exclude a 1.2, 1.3 as being a donor of sperm in that sample, would they?
There is an error. There is a 1.3 allele that you didn't find. That is significant.
In your view are you as qualified as he to render opinions about matters like this involving sexual assault samples?
Your Honor, I have one more record I would like to be marked as a People's exhibit.
It is sort of like if you went into your house and you noticed a cockroach and then the next day you noticed another one, it would be an indication to you that maybe there are other roaches around and you maybe better do something about it.
So you made the decision to change the title 'Contamination incident' and change those two words of the title to 'Unexpected alleles'; is that correct? That was to make your chart more accurate, correct? And that is because there may not be contamination in many of the instances that you describe in this chart, correct?
It is a little confusing, isn't it? I guess, yes, that is fair.
In your view is it important to have experience in actually using a technique to be able to interpret it, as you have done in this case? No.
How many other laboratories have you examined all of their PCR DQ-Alpha typing strips for a 15-month time period? For that long a period, none — no other laboratory for that long.