All right. Thank you, ladies and gentlemen. Please be seated. Let the record reflect that all the jurors have now rejoined us. Dr. John Gerdes is on the witness stand undergoing direct examination by Mr. Scheck. Mr. Scheck, you may continue.

Thank you, your Honor.

Dr. Gerdes, let's turn to the issue of error rates. How are errors known? How do you learn about errors in the clinical context?

Well, one way to determine if an error has been made in a clinical lab is by and through the proficiency testing. The NMBB program, as I explained earlier, that is blind proficiency testing and what that means is every week if we have sixty samples that are sent to us to be typed, ten of them are incorporated as controls and we don't know which ten those are, and then if during the course of our typing if we make errors, that will pick up the error immediately.

Besides proficiency testing, which we will discuss a little bit more in a moment, are there other ways that a clinical laboratory finds out, such as your own and others that you are familiar with, that it made a mistake in error?

Again, in the case of bone marrow transplantation, as I explained, if we give marrow to a recipient that doesn't match, the patient comes down with graph versus host disease and can die because there was an incorrect match, so in the clinical setting mistake is fairly drastic.

So in other words, would it be fair to say that in that in the clinical setting you do a DNA typing bone marrow solid organ transplant and post-transplant infectious disease screening, and if your laboratory made a mistake, that would be become known because the--because of an actual outcome of the patient and you would hear about it?

That's correct.

And you would really hear about it, wouldn't you?

Yes, yes.

Now, in the context of a forensic test, is there a difference in terms of finding out whether or not an error has been made in a case or how easy is that to determine?

Well, put it another way: In terms of the forensic labs, as opposed to the clinical labs, there is no independent outcome, objective independent outcome that you can make a comparison to? Would that be a fair statement?

That's correct, yes.

So would you agree that in respect the error rates are not as well known in the forensic setting?

That is true.

In terms of your examination, the laboratory standard in clinical laboratories in terms of the protocol, procedures and the standards set out by accrediting agencies compared to the 23 forensic labs that you've looked at, which has the higher laboratory standards in terms of specimen handling and basic fundamental DNA techniques?

In a clinical lab we have, as I mentioned, a number of agencies that inspect us. We are always being inspected, and these agencies, they have mandatory guidelines, mandatory requirements, there are laws that we have to follow, and so the standards are fairly high and they are mandated to be high by law. In a forensic laboratory there isn't a similar organization that would--that has mandatory accreditation for those kind of laboratories, so they are lower standards in the forensic lab.

Now, the Asclad program, to your understanding is that--that is a voluntary program?

It is voluntary.

And I think there has been testimony in this case that DOJ and Cellmark are but that the LAPD is not. Is that your understanding?

That's correct.

All right. Now, in terms of proficiency testing, you began to tell us before, could you tell us briefly a little bit more about that national bone marrow donor transplant program. How many samples does your lab get a month?

At the present time we are getting 200 actual samples a month of which forty are controls that are incorporated into what comes into the lab and we don't know which forty are the controls.

So you think that the--all sixty samples that you are getting that month that you have to do DNA typing on, you think that is a regular case?

They are not marked in any way. We know ten of them are controls, but we don't know which ten.

And in terms of the definitions that we have been using in this case, would you--is that an external blind proficiency test, that is, it is done by an outside agency and the laboratory is blind, that is to say, you don't know whether it is a real case or not?

That's true.

All right. And the--you also undergo what we've called open proficiency tests, that you know you are being tested and you are sent samples; is that correct?

That's correct.

And have you reviewed the proficiency tests that have been done by LAPD, DOJ and Cellmark and other forensic laboratories?

I have.

And in terms of the proficiency tests themselves, the kind of samples received and the nature of it, in your judgment which is more rigorous, the clinical proficiency tests you have described or the forensic proficiency tests that you've reviewed of LAPD, Cellmark, et cetera?

Well, in order for a proficiency test to be realistic, it should really mimic exactly what you are trying to test as far as the ability for a lab to do something, and so what you attempt to do is design these samples to be exactly like what you are claiming to test. And in a clinical setting that means that the samples that are sent to us are essentially identical to a patient sample and it comes in and it is incorporated into our normal run for that day, our normal testing for that day, and then it is treated in the same way and you report the result, just like you would any patient. In the forensic setting that is a little more difficult because of the fact that, remember, these samples come in from an infinite variety of states, conditions, places where they were found, how old they are, what they were exposed to in terms of humidity and temperature and length of time, and so it is extremely difficult to mimic that on all different kinds of varieties of specimens that you might do. And mixtures are another problem. You should really do the hardest possible type of specimen to get an idea about the maximum error rate when you have the hardest test.

Well, Dr. Gerdes--

That would be a degraded mixed specimen with very little DNA, for instance.

Have those kind of--are those kind of samples done in the proficiency tests that were performed by LAPD, Cellmark and DOJ?

Objection, no foundation.

Sustained.

Have you reviewed the CTS and CAP proficiency tests that were done by LAPD in this case?

Yes.

Have you reviewed some that were done by DOJ?

Yes.

Have you reviewed some of Cellmark's tests?

Yes.

All right. In your judgment, in terms of the nature of proficiency tests, which are more rigorous, the clinical or the forensic?

Same objection.

Sustained.

In terms of the samples that are given to the laboratories for proficiency tests, which are more rigorous, the ones encountered in the clinical work or the forensic?

Same objection.

Counsel, the problem was the testimony was the same or some of the proficiency tests. I don't know what range. He may have looked at one proficiency test per lab. That doesn't tell me anything.

Oh.

Have you looked at every proficiency test done by the LAPD?

Yes.

How many proficiency tests did you look at from the Department of Justice?

Umm, I can't present a precise number. It was somewhere in the range of between six and ten.

All right. Incidentally, are these the--the Department of Justice and LAPD doing some of the same tests?

Yes.

In other words, the CTS and CAP tests that they were doing are precisely the same sample?

But sent to different labs, yes.

Okay. Now, in terms of what you reviewed in those laboratories, and comparing that to the clinical tests you are familiar with, which would you say in terms of the samples are the more rigorous ones, more rigorous proficiency tests?

Same objection.

Overruled.

In the proficiency that I saw, all of those proficiency tests involved unmixed blood specimens from known individuals or easier types of specimens, not--none of them were degraded or mixtures, for instance.

So would you say the forensics?

I would say forensics.

Excuse me, gentleman. You are going to have to stop talking at the same time, especially you, Mr. Scheck. Slow down a little.

My apologies.

The court reporter is writing me messages.

Finally, Dr. Gerdes, are you familiar with the use of statistics in terms of DNA tests?

Yes.

Are you an expert in population genetics?

No.

Are you a user of some of those statistics?

Yes.

Are you aware of statistical controversies in terms of--withdrawn. Let me put it this way: To your knowledge, in terms of the DNA test results in the clinical setting, is there any controversy over statistics?

There is.

In the clinical setting?

Oh, in the clinical setting, I'm sorry. For the majority of the work we do in that setting there is no statistics involved.

When you say no statistics involved, what do you mean?

Well, the kind of questions you would ask, for instance, if we are doing HLA, we are asking--the question is we have two individuals, do they have precisely the same HLA? And you have the two individuals right there, so you don't have to calculate what chance at random that would happen in a population because you know what two individuals you are looking at. You are either looking at--you are looking at a known person who needs that transplant and you are looking at a donor and you know both of those people.

You are aware in forensics, I take it, one uses databases to calculate statistics?

Yes.

Are you aware of the statistical controversy in forensics?

Objection. I'm sorry. No foundation, calls for hearsay, also assumes facts not in evidence.

You can say yes or no he is aware of it.

Are you aware of it?

I am aware of it.

Incidentally, doctor, have you reviewed the report on "DNA technology in forensic science" by the national research council?

Yes.

Do you regard that as an authoritative text?

Yes.

Do you rely upon the conclusions in that text?

Yes.

And have you read the section there dealing with statistical controversy?

Yes.

Just the fact--all right. Your Honor, I believe we are finished with this board.

Now, Dr. Gerdes, did you conduct an examination of DQ-Alpha hybridization strips at the Los Angeles Police Department that began on May 20, 1993, strips from May 20, 1993, through August 25, 1994?

I did.

And could you please tell us what kind of data these strips came from?

Excuse me. Objection, no foundation.

Overruled.

Well, did you get these from the Prosecution?

Yes, they were provided.

Did you go to the lab and look at some of them?

I did.

All right. Were you provided something that were known as LAPD validation studies?

Yes.

What were those?

Well, LAPD, when they first set this up in May of 1993, they collected specimens from--

I'm sorry, excuse me. Objection, no foundation.

Sustained.

All right. Did you receive documentation from the Los Angeles Police Department entitled "Validation studies"?

There is a cover page that discusses that they looked at various specimens for the purpose of validation. I don't think it was really titled "Validation studies," but I have called it that and they use it as--and on that page it discusses using those specimens for the purpose of validation.

Your Honor, I would show the witness what has previously been marked as Defense 1181-A.

All right. Have you shown that to Mr. Clarke?

Actually, while we are at it, 1181-B as well.

All right.

You said Clarke and I thought one Clark was enough.

Clarke and Clark. Thank you.

Yes.

And I will just briefly put those on the elmo to remind the jury what we are looking at.

Could we--

So this is a document entitled "PCR method validation training record"?

Yes. I didn't recall what they actually titled it as such. It looks like they did.

All right. And--and 1181-B is--represents what?

This is a list of some of the standards, the individuals that they use for the validation.

So in other words, what they are indicating here is that a series of bloodstains, saliva from swabs and cigarette butts, hair, blood saliva and hair sets, mock vaginal swabs, nine known bloodstains, family studies, these are known samples that were then sent out to the analysts to type; is that correct?

That's correct.

And it indicates here that all the above validation work was performed by Erin Riley and Collin Yamauchi and every validation sample either gave the expected typing result or no typing result was observed--at no time was an incorrect typing result observed?

That is what it says.

All right. Now, did you review all the hybridization strips based on this PCR validation set of samples from the LAPD?

I did.

All right. And did you review as well the--all the LAPD proficiency test strips from their internal tests, from the proficiency tests, from the collaborative training service and the College of American Pathology?

Yes.

I'm sorry, objection; no foundation.

Overruled.

Did you review samples that were received by LAPD from what is called their Korean database?

Yes.

All right. And that is blood samples from people that are self-described to be Koreans that they typed; is that right?

That's correct.

Did you look at case work strips during this period of May, 1993, to August, 1994?

There were some case work strips, yes.

All right. Did you look at as many as you were permitted to see?

I looked at everything they gave me.

All right. And when you evaluated the case work strips--withdrawn. On the proficiency tests, the validation studies and the Korean database samples, did you know what the sources of those samples were?

Yes.

So they are known types; is that correct?

They are standards, that's correct.

Standards. Now, in the case work, when you looked at the various strips from case work, did you know the source of every one of those samples in a case work--set of case work strips?

No.

Which--which strips would you know in a case work in a case work sample came from a known source?

That would be the positive control and the negative controls from that particular case and any sample that was referred to as a reference sample would have been known to have come or defined to have been derived from one individual.

All right. Let's just review it now for a second. The positive control, is that the 1.1/4 DNA sample that comes with the DQ-Alpha kit?

That's correct.

The negative control or the negative controls, there is two kind of negative controls?

Yes.

All right. One kind is the--what is known as the extraction control, that would be a sample that is not supposed to contain DNA?

Yeah. The extraction control consists of basically using the substrate which would be a control swab or a piece of cloth that is run through the DNA extraction process all the way through to the typing process and it controls for foreign DNA that might have been incorporated or accidentally introduced into that test during those procedures.

And is there another control that is introduced at the end of the process when you amplify up the DNA?

Yes.

That is called, what, a negative amplification control?

Amplification blank. Some people call it a water control. It basically does not go through the procedures involved in extracting DNA. It is incorporated at the stage where you copy the DNA or amplify the DNA, and so it only controls for the accidental incorporation of DNA at that stage.

Now--a known reference sample in a case would be, let's say, in a sexual assault case if they took a sample from the victim, that would be a blood sample that would be considered a known; is that right?

It is considered to have come from one individual. I wouldn't know the type--the anticipated type, but it is--I think I--it is safe to assume that that is defined to have been obtained from a single individual.

So--

It should not be a mixture.

So when you are looking at a DQ-Alpha strip from a reference sample in case work from a known individual, you should see no more than two alleles or two--

That's correct.

If you see three alleles--

That is an indication that it has to be a mixture, or in this case, since it was defined as having come from one individual, if you have an indication of three there, then that has to be contamination, that has to be foreign DNA that was incorporated or somehow got into that sample.

All right. So when you looked at the case work strips from LAPD, you were examining to see whether there was contamination?

Yes.

And you would determine that by looking at the positive controls that come with the kit, known reference samples and the two kind of--and the negative controls?

Correct.

All right. And incidentally, if you get a dot, something showing up in the negative control, does that indicate contamination?

Absolutely.

All right. So did you conduct an analysis of all these strips from the validation studies, the proficiency tests, the Korean database and all the case work samples that you could see at the LAPD from May of 1993 when they started, through August of 1994?

I did.

All right. And what I would like to do now, your Honor, is put up a chart and will--doctor, do you have a set of strips that would illustrate how you--how you read them and went through your analysis to determine contamination?

Yes.

All right. And I have shown this previously to Mr. Clarke. Could we mark this strip Defendant's next in order?

1286, Mrs. Robertson? 1286.

While Mr. Clarke is looking at that, I would like to pull out a board.

All right. We will mark this chart next in order, 1287.

This is called "Strips: Percent of contamination and/or artifacts." Mr. Scheck.

Yes. First, maybe we could pull up and look at the entire sheet here and you can describe for us what this is generally.

Now, is this the way you--is this the form in which you received most of this data?

Yes.

All right. And up on the top it says what, "DNA hybridization record"?

Yes.

And that would indicate--8/25, does that indicate the date that this hybridization was run?

Yes.

And on top it indicates an analyst there. Is that a gentleman that you know works in the DNA lab whose name has been discussed?

Yes.

Harry Klann?

Yes.

Underneath it where it indicates "Confirming analyst," that would be the initials of who?

Collin Yamauchi.

All right. And below that are the strips; is that correct?

That's correct.

And below that are--is the data that the analyst would write down recording what the analyst is seeing or concluding; is that correct?

That's correct.

It least what each of those is, the negative control, the positive control and the samples?

Correct.

All right. Now, if we could focus in here, did you find contamination in this series of strips from the Korean database?

I did.

Excuse me. Objection, no foundation.

Sustained.

All right.

I'm sorry, move to strike the answer.

The answer is stricken. The jury is to disregard.

Would you describe for us what you observed in these samples and what conclusion you drew?

Same objection as to foundation.

That is a multiple question. He can describe what he observed at this point.

Please first tell us what you observed.

I observed the typing results of these particular strips.

All right. And what were these based on--what were these--what were these strips to represent?

In this particular case they were typing the Korean database. The top strip is the negative control.

So--

That is the amplification negative control we talked about earlier. The second is the positive control which is the positive control that is incorporated in the kit. That is the 1.1, 4 control.

Excuse me for a second, Dr. Gerdes. Maybe it would be easier, I think that--

Do you want me to go up--

You have learned how to use this device?

I think so.

Your Honor, may he leave the witness stand and use the pointer, please.

Yes.

Can you please describe for the jury how you analyzed this strip?

I'm sorry, your Honor. I have an objection as to foundation as to this entire exhibit.

Sustained.

What--reviewing these records--

What do these records represent? What do these strips represent based on these records?

Same objection.

Sustained.

Where did you get this from?

I obtained this from--during discovery it was obtained from the Los Angeles Police Department.

And based on these records, what were you told this represented?

Objection, calls for hearsay.

Sustained.

All right.

From whom did you learn what this was?

Well, same objection.

Overruled.

Did you have any conversations with Mr.--did you ever visit the lab and discuss how these records were put together with Mr. Yamauchi and Mr. Matheson?

Yes.

All right. And based on those conversations what is your understanding of what these records represent?

Same objection.

All right. I'm going to allow testimony on this subject to a foundation being laid for what it is.

All right. Proceed.

My understanding is that this represents a particular typing run of--on which there are a number of individuals from the Korean database and a positive control and a negative control.

All right. Could you please now in analyzing the strips--back up--okay. Could you please--

Sorry. I still have the same objection as to foundation.

Noted. Thank you. There will be a standing objection to this type of document.

Please tell us what these--what these strips are, according to their records.

The top strip--and I'm trying to get an arrow to work.

Draw on that and see if you can move it.

He has got it.

Okay.

Move it around now.

Okay. This particular strip right here, (Indicating), that moves along all the way across, this strip.

Appears to be the strip marked 261-1.

What is that?

Correct. 261-1 represents the negative amplification control for this particular set of strips. Whoops.

What is the next one down?

The next one down here represents the positive control, 261-2, and as you can see, it types as a 1.1, 4.

Okay. What is the next one down?

The next series represents different individuals that were in the Korean database.

All right. And how many in--could you look--directing your attention first to 261-4, how many alleles was that type to have?

I have to look at the sheet, but I believe they typed this as a 1.1, 1.1 because of this dot here, (Indicating), which represents the 1.1.

Can we move out a little on this.

And this dot here which represents the 1, so it was typed at a 1.1, 1.1.

Okay. Just move it up. Pull it back. Pull it over. Could you move to the next one.

Please describe on any of these known from the Korean database did you see more than one genotype or extra allele?

Yes.

Could you tell us which one?

Well, none of these particular samples were typed as having a 4 allele and you can see here, and if you bring it up a little further I think it is clear, that you can see a 4 dot here, (Indicating), confirmed by the 1.2, 1.3, 4 dot here, (Indicating), and that is on this particular sample and it is also on this particular sample here, (Indicating), and it is on this particular sample here, (Indicating). In this case there is a dot present because there was a 1.2.

Now, what does the existence of--what does the existence of the 4 lighting up at 261-4 and the 1.3 dot lighting up at 261-4 on that strip and the 4 and the 1.3--the 1.2, 3, 4 dot lighting up on the 261-6 strip? What is the significance of that?

Well, the significance of those weak dots are that there really should not be more than two alleles here. That represents a third allele. These samples came from a database presumably from single individuals, and therefore that represents human DNA that shouldn't be there, and that is what our definition of contamination is.

And that holds I take it also for 261-7?

Yes.

All right. Now, if we move back a little bit, focusing on another part of this sheet.

How did LAPD--what observations did they make with respect to some of these samples in terms of that 4?

They recorded the presence of those, but didn't incorporate them into their typing result.

So in other words, in the fourth strip down that you are pointing that they recorded as a 1.1 with a 4; is that correct?

Correct.

I'm sorry, your Honor, would it be possible to have the chart moved?

Yes.

Thank you.

Put this down for the time being.

And in the seventh strip down, again those--what does that nomenclature indicate "4 less than c"?

It indicates that there was a 4 dot observed, but it was less than the control dot and the control dot, if you remember, is the dot that determines whether or not there was an adequate amount of DNA to proceed with typing.

Now, in your opinion does the fact that that 4 is less than the control dot, does that mean that the--those 4's are not contaminants?

Absolutely not.

Why?

Because of--there are a number of possible explanations as to why you would have that additional dot, and there are no known descriptions of that particular dot, the 4 dot, having what are known as cross-hybridization problems, and so the only explanation for that particular dot is that it is real, not an artifact, and that that dot represents additional human DNA.

So in other words, additional human DNA, that 4 dot means that those samples are contaminated in some fashion somehow?

It does.

Now, did you, in going through the strips that you looked at--and I would ask that this be marked as--what's next in order?

1288.

1288.

Your Honor, if I make a request is that--to expedite this, could we reconfigure the errors on the typing strip at the break and then print it out?

Yes, yes.

Should we give that a number and I will make sure to do that.

Yes. It should be 1286-A.

We will do that with Mr. Clarke at the break. And so this would be 1288.

Now, did you--did you conduct a--

Did you prepare a chart indicating in all the samples you looked at that the LAPD from the validation studies, the Korean database, the proficiency tests and the case work, all the strips you looked at, how many times you saw a 4 allele in a known sample where it shouldn't be which in your opinion was contamination?

I did.

All right. And could you please explain what this chart represents?

Well, this represents a graphical picture of that beginning in May. What we see is that they really didn't experience it initially. I didn't observe, anyway, these extra 4 dots--

Well, before you leave there, just so we understand what we are talking about, when that box says 5/93 and underneath it 0/32, what does that represent?

This is May of 1993 and under here are the number of type strips I looked at and how many times I found that 4 as an additional dot.

All right. So--

And then represents--the percentage would be calculated from this and then displayed.

So if I could move through it quickly then, 5/93?

Nothing.

32 strips, nothing?

6/93, 50 strips, nothing; 7/93, 56 strips, nothing; 8/93, 1 out of 38; and 9/93 now, 9 out of 136; and 10/93 there were 5 out of 97; and in 11/93 there were 2 out of 11; in 12/93, zero out of 17; January of `94, 18 out of 50; February, zero out of 8; March, 1 out of 5; April, 18 out of 46; May, 4 out of 45; June, 1 out of the 16; July, 1 out of 12; and August, 10 out of 61.

All right. Then if we pull back and look at the chart as a whole, you then looked at those absolute numbers and created a sample bar graph of what that--what this 24 allele contaminants represented?

Correct.

As a percent?

Correct.

All right. And as far as the 4 allele is concerned, when you see an extra 4 allele, there is no question in your mind that is a contaminant?

It has to be contamination.

Now, I would like to look at what I would ask to be marked as--1288, this will be 1289.

All right. 1289.

Did you do a similar analysis for the 1.2 allele?

Yes.

All right. Now, what is the 1.2 allele in the strip? Is there any particular dot for the 1.2 allele?

No. There isn't a specific probe for this particular allele. You depend upon the 1.2, 1.3, 4, dot and interpret it in context of the other dots to make a decision as to whether it is really a 1.2 or not.

So did you perform a similar analysis in terms of numbers of strips and percentages whenever you saw an additional 1.2 allele on a known sample?

Yes.

All right. And this table represents that?

Yes.

All right. And whenever you see a 1.2 allele, an extra 1.2 allele, in your opinion is that definitely contamination?

It is definitely contamination.

Now, I would ask to mark what is 1290.

Did you do a similar analysis for the no. 2 allele?

I did.

And does this chart represent that?

It does.

All right. And it is a percentage analysis as well as the absolute analysis?

That's correct.

Just to move in, just to give us a sense of it, for example, in January of 1994, you are seeing what there?

20 out of 50 strips which whatever percentage that is, and in February there was zero out of 8; 2 out of 5 in March; 17 out of 46 in April; 2 out of 45 in May.

Et cetera?

Et cetera.

All right. Now, I would ask to mark this as Defense 1291.

So marked.

This would be an analysis of the 3 allele?

Correct.

And I take it that--

They don't seem to have too much of a problem of the 3 allele. There was only one observation that was made, but I can't read it.

3 out of 61?

3 out of 61, yes.

When you see a 3 allele is there any doubt in your mind an extra 3 allele, that that is contamination?

It is definitely contamination.

Now, are there certain--what is an artifact?

In this DQ-Alpha system?

It is a result, a mistaken result that is as a result of a flaw in the system.

All right. Now, is there some--in other words, you see a dot that is a result of some defect in the system and it is not necessarily--

Correct. A limitation of the typing system itself has created that signal so that you can't really determine if it is real or not.

Mr. Scheck and Dr. Gerdes, please, you can't talk at the same time.

All right.

Is there a--with the 1.1 allele, is there sometimes something that arises as an artifact?

Yes.

All right. Do you have something from the user guide that can demonstrate that for us?

I believe so.

I gave it to you.

It is a trick question.

I ask that this be marked as--

1291.

1291.

Excuse me, 1292. 1292.

1292.

Can we focus in on the strips known as 7 and 8 there?

Now, what does this illustrate, doctor?

Can you focus that a little better and get it enlarged a little better? These two strips represent a typing in which you have a classic example of what is known as the DX gene and that is an artifact in this particular typing system. It is a limitation of this system and it occurs when you have alleles other than the 1, so here we have a 2 and a 3, for instance, in both cases, and the C dot is found and there is you can barely see it, but there is a 1.1 here in both of these strips.

Now, is that one--

Here, (Indicating), and here, (Indicating).

All right. Okay. All right. Now, is that--maybe we should have Mr. Harris do this.

I'm sorry.

Okay. Why don't you put another dot on the other 1.1 there. Okay, 1.1 there. And could you write on the upper left-hand side of the top of the chart, "DX." Do we know how to do that?

I believe so. (Witness complies.)

All right. Now, the--I take it then when you see a light 1.1 dot but no 1 dot on the far left-hand side--is that correct?

That's correct.

--that is what you call a classic DX?

Yes.

And what is--if you can very, very simply, just tell us what that is? Why does that happen?

Well, the explanation is that there is another gene that is similar enough to the DQ-Alpha gene to have some of the probe signal to give some probe signal on that particular dot, and so if you have a lot of DNA, and by the literature that means greater than six nanograms of DNA, you can sometimes see this artifact.

Now, let's assume that in these two instances that we are looking at where the arrows are pointing, those are the expected genotypes from that sample, is a 2 and a 3, right?

Correct.

Because the 2 and the 3 dots are lighting up?

That's correct.

Let's assume for--that the real genotype that you are typing is a 1.3, 2 or some other allele, a 1.3 or a 1, that would light up the 1 dot on the left-hand side.

Okay.

And you also saw a 1.1 dot.

Yes.

Could you tell whether that was the DX or a contaminant?

No.

You couldn't determine either way?

Because the definition of DX requires that you have actually two probes here, one, the 1 dot confirms that there is actually a 1 there, so if there is another allele there that is a 1, that dot will light up and now the 1.1 is really confirmed by that first dot and so you can't make a decision, a scientifically sound decision, unless you do additional testing, sequence it or do something else, because in that particular set-up it could be a DX or it could be a real allele.

Well, is it a good thing in terms of a typing systems to have this kind of possible DX artifact that confuses interpretation?

No. The problem with this is if it is from one individual, you can sometimes, as in this case, make an excuse that that 1.1 is just an artifact, but if it is in a forensic sample and you don't know it is from one person, now you can no longer decide is that real or is that due to a mixture, and especially in a situation where you have that 1 dot.

Now, when you went through the DOJ typing strips, did you do an analysis of instances where you found extra 1.1 alleles?

I did.

Your Honor, could we print this out from the elmo now?

Yes.

1292-A.

So marked.

And at an appropriate time, your Honor, I would ask to pass it to the jury because I understand that it is in terms of seeing the light dots it is easier on this one than it is on the monitor.

I don't know. Can you see it right now?

Sorry?

It is clear on what the elmo has printed out there.

Yeah. I think you can see it on what the elmo has printed out.

All right. Let's do it now since if we do it later the jury may not recollect what this relates to.

Thank you.

All right.

Dr. Gerdes, you--would it be pair to describe--

Excuse me. I'm sorry, counsel. The record should reflect that each member of the jury has had an opportunity to view 1292-A.

And 1292.

-a.

Now, 1292-A can be described as the classic DX, right?

That is the classic example that is used to show this artifact phenomena, yes.

How many times, in your analysis of looking at extra 1.1 alleles, did you see the classic DX without a 1 dot?

Of the alleles that were counted as 1.1 extra dots, 7.6 percent of those were due to this classic kind of example.

And would that be 9 out of 46?

Yes.

All right. Now, what is the next one, your Honor, I'm sorry?

1293, I believe.

1293.

Yes, 1293.

Did you chart in the same manner that you did--did you create a chart of extra 1.1 dots that you saw?

Yes.

And is this that chart?

It is.

This again is representing May through August of 1994?

Correct.

So what you are telling us is that nine of these 1.1's were the classic DX?

Correct.

And the rest of them were 1.1's where there were 1 dots present?

Correct.

Where it could be a contaminant?

Correct.

And we would call the classic DX an artifact?

Correct.

All right. Now, with respect to the 1.3 dot, did you compile a similar chart of your observations of all the strips?

I did.

And I would ask to mark this 1294.

Now, the 1.3 allele, when you see an extra 1.3 allele, are there situations where that could be an artifact as opposed to a contaminant?

Yes. It has been observed and described in the literature that on this particular allele, if you use DNA concentrations greater than six nanograms, you can see faint signals on that particular dot.

Is it your understanding that the samples in question were contemplated to have as much as six nanograms, the ones that you looked at?

Some of these may have, yes.

All right. Does this represent the number of 1.3 alleles that you saw?

It does.

When you are looking at a forensic case and you see a 1.3 allele, an extra 1 dot, and the sample has less than six nanograms of template DNA, what is the best interpretation of that?

If it is less than that amount of DNA, the best interpretation is that it is a true contaminant.

All right.

It is extra DNA.

Now, did you--now, I would like to turn to the larger chart. Did you do a compilation of all those individual allele charts, pulling them all together to look at all the strips that you examined? And this is--what did we mark this as?

1287.

1287.

On 1287, did you do an overall chart indicating what you found in terms of contamination and the artifacts you described, the--the seven DXs?

Yes.

Over this period between May of 1993 and August of 1994?

I did.

And this represents absolute numbers and percentages?

Correct.

Now, did you also look at what are known as runs?

Yes.

What is a run as opposed to these strips?

On a given day you will test a series of strips. It can be a minimum of perhaps eight or as many as thirty or forty, but on a given day, if you look at all of those strips, that is called a run.

All right. Now, did you do--so in other words, these are the individual strips by month?

Those are the individual strips.

And did you compile an analysis of runs, all the strips in a day, to see whether or not there was a definite contaminant on a run that didn't represent an artifact in any way?

I did.

All right. I would like to take a look at that.

Can you please describe for the jury what this chart represents.

This represents--

I have to mark this. This would be 12--

1295.

1295. And there would be a chart entitled "Runs: Percent with contamination by month, May, 1993, through August of 1994."

Correct.

All right. Would you please describe what this represents.

Well, it is titled "Percent with contamination" now, not "Contamination and/or artifact," so what I would do is if you look at all the strips that are done on a given day, if you--you can look at those in the context of one another and make a scientific decision as to whether or not this is true contamination or not. And the kind of things that would convince you that it is contamination is, let's say, for instance, I found a weak 1.1 on one strip that might be DX, might be an artifact, but if I also found a 1 dot on the no DNA control on that same day and I found a 1 dot on the extraction control on that same day, and I found that same 1.1 dot on other strips that had 1's where I can't make a decision, if you look at that all in context and look at the whole pattern, it is--you can confirm that this is this is not just a random thing that happens at a low percentage of the time due to this DX artifact, it is really contamination.

Well, these charts represent--let's just take in May when--in May of 1993 there were two runs but you didn't see any contamination in that run in that laboratory?

That's correct. They started in May and they did 45 strips, if I remember correctly. None of those strips showed any indication of the kind of things we are talking about as either artifacts or contamination.

All right. And then--and that represents two runs?

That is only two days, that's right.

Then all of a sudden June of 1993 what happens?

And we are talking about this is contamination negative control or the 4 allele, 3 allele, 1.2 allele and the 1.1 under circumstances where you can confirm it is definitely contamination and not an artifact?

That's correct, by the criteria I mentioned earlier where I can definitely confirm that this is contamination, not one of these artifacts.

Now, as a DNA laboratory director, when you start seeing on the runs in June of 1993 that fifty percent--what would that be, fifty percent--

Fifty percent.

Fifty percent of contamination, as a laboratory director what do you do when you are dealing with the PCR system?

If you don't do that and you just ignore the contamination, is there a problem because this is chronic and accumulates?

Yes. If you don't identify the source of the contamination and remove it, it stays in the laboratory and you don't know where it is and it is going to come--it will become evident on a sporadic and random basis initially, but eventually it will tend to build up. And I didn't point it out on those other strips, but you can see a build-up from `93 to `94 in the occurrences. You can see it here as well on the run strip, that you started with nothing, all of a sudden they have 50, now they are up to 66 percent or two-thirds of the strips and on some months--

Two/thirds of the runs?

And on some months it goes all the way up so that all of the runs are contaminated.

Now, when you find contamination like this, it could be from any of a number of sources or procedures?

Yes.

And what you are saying, that should laboratories document contamination?

In my opinion it is absolutely imperative.

Should you document the steps that are taken to correct the contamination problem once you have it in your laboratory?

Yes.

How important is that?

I think it is extremely important. In fact, there are forensic guidelines that state that that is a recommendation that they should do, that they should actually do that in forensic labs. We do it in research.

Stop right there. You say forensic guidelines. What are you talking about?

It is called a Twgdam guideline and it stands for the technical working group of DNA analysis methods.

All right. I mean, it is just basic sound laboratory practice that if you detect the amount of contamination that you saw on the runs at the LAPD right away, that you have to document it and document all the steps that ought to be taken to correct it?

I should document that it happened, document what you did to fix it, document what you did to determine it has been removed from the lab and then proceed.

And could this contamination also come simply from the methods that you use to handle samples?

Excuse me. Objection, calls for speculation.

Sustained.

All right. Does contamination arise from sample handling methods?

Objection, no foundation.

Sustained.

All right. In your experience as a DNA laboratory director, what are the various sources of contamination problems such as the ones you saw here at the LAPD?

Well, as I have described that already, you can have introduction of foreign DNA by handling a specimen in such a way that if you are not careful that you can get transfer of DNA from one sample to another. So if you are working in this case with these samples are known to come from single individuals, that is one possible explanation as to how it got there, is you have--you handled it, you know, in a sloppy manner and you accidentally introduced somebody's DNA into that particular sample during the process of preparing it and handling it.

Now, just in terms of the strips of the samples that you were looking at, you were looking at known samples from those validation studies; is that correct?

Yes.

You were looking at samples as part of a database where the laboratory receives a single blood sample and has to type it?

Correct.

You were looking at the positive controls on case work, correct?

Correct.

Now, those are not supposed to be, I take it, degraded samples or mixtures?

No. Those are going to be the easiest possible sample to type. You've got a lot of DNA there or adequate DNA certainly and they should be defined from single individuals.

So the analysis of contamination that you've just presented to us is based upon known exemplars or easy samples?

Yes.

Your Honor, actually I think it is probably a good--before I get to the May through July chart--