Dr. Cotton, when you referred to the phenomenon in your laboratory where there was multiple instances of negative control failure occurring, when did that happen? Do you know the date?
No, I don't know the date without--it was sometime in the last two years, but I don't know the exact date.
And for each of those four or six cases, how many samples were there tested where the negative controls failed?
I have--I have absolutely no idea. It wouldn't been--wouldn't have been necessarily all the samples in that case. It could have been like one sample that was done at that time. But how many, I don't have any idea.
Is there a centralized log which one could access to see how many samples had the negative controls fail in that period?
Would you agree, Dr. Cotton, that it is important to find out the source of the contamination causing the negative controls to fail so that you can correct the cause and thus reduce the likelihood of reoccurrence?
It's nice if you can find out the source, but many times, that's not possible. So I would not agree that it was essential or necessary.
Well, in this case, Dr. Cotton, isn't it true that a reagent blank that was supposed to be blank displayed DNA dots?
Do you want to rephrase that, "this case"? When you say "this case," are we talking about the--
In this case, right here in this courtroom involving Mr. Simpson, isn't it true that a reagent blank that was supposed to be blank displayed DNA dots?
And I believe you said on direct examination that there can be two possible explanations for that occurrence. Do you recall that?
And I believe you said that one possible explanation is that there was a minute contamination during the original extraction which was magnified when it was concentrated along with item 7. do you recall saying something like that?
Yes, I do. That's--that's a good restatement of what I said or what I meant anyway.
Now, is it also possible that a second explanation would be that during the concentration procedure itself, the contamination occurred?
Now, would you agree, Dr. Cotton, that if one doesn't know which explanation is correct, that to be cautious, it is incumbent on the laboratory to assume that the contamination occurred during the initial extraction?
No. I mean--there's two possibilities. There's no way to distinguish between them. So I think it would be inappropriate to assume one or the other.
Okay. Well, if there's no way that you can prove which one it is, then certainly the explanation that the contamination that led to the reagent blank failure occurred furthest back in time as a possible theory; isn't that correct?
All right. Now, the other--there were other samples that were extracted during that extraction when you set up the reagent blank; isn't that right?
And would you agree that the other items in this case that were extracted at the same time as the commencement of the reagent blank were items 78, 56, 52, 47 and 12?
The reason I'm having trouble is that our numbers are different than the item 78 number. So what's in the information that I'm looking at are just our own numbers and it doesn't say item 78. So we have to--let me just find the piece of paper and take it out that allows me to easily go back and forth between the two sets.
Okay. And, Dr. Cotton, given the presence of DNA on the reagent blank negative control happening when you ran that batch of items that you just enumerated, shouldn't you have gone back and repeated the extraction procedure for those five items in the case?
Doesn't your protocol, Dr. Cotton, say that if the reagent blank fails, then all the samples must be reextracted?
The reagent blank for the other samples apart from our samples 06, which is item 7, when that reagent blank was originally typed, it typed with no DNA signal visible at all. Therefore, the sets of--the samples apart from sample 7 that did not get put through the microcon could not have been effec--even if we assume--if we just say for the sake of assumption that there was some minor level of contamination that got concentrated later by the microcon, it was not at a sufficiently high level to produce any product whatsoever, and, therefore, it couldn't have affected the types that we came up with on the set of samples apart from the one that went through the microcon.
Well, Dr. Cotton, isn't it true that a contaminant does not necessarily have to contaminate all samples in the same quantity?
Given that all these things are in solution and based on the solutions are--would be normally considered to be even distributions of molecules, I don't really think that that's an important consideration.
But isn't that true, Dr. Cotton, only if you can assume what the cause of the contamination of the reagent blank is?
No. I just--I think that since the reagent blank contamination clearly didn't show up in that first set of samples, that it does not present any problem to that first set of samples.
But, Dr. Cotton, doesn't your protocol explicitly state that if the reagent blank fails, then the entire experiment for the samples that are run with it should be deemed inconclusive? Isn't that in your protocol?
I'd have to look at the protocol to see what it specifically states, but in the protocol, we don't usually use the word "experiment." So would be better to go back and look at what it specifically states. I believe that it states that if possible, the sample should be run again.
And does it also state to your recollection that the--that the data should be deemed inconclusive?
Yes, it states that, but I can't tell you right off the top of my head if it states that as an absolute or it should be--or that that's a possibility that should be considered.
KEY QUOTELet me just show you the following document, see if it refreshes your recollection on this particular point.
Thank you. Doesn't your protocol say that the--that when this happens, the test will be considered inconclusive? When I say "this," I mean when the reagent blank fails.
It doesn't say it could possibly be considered inconclusive, does it? Doesn't qualify it that way?
Well, there is a sentence later on that refers to the Ph.D.s using their discretion when--and I can't restate the sentence exactly. Maybe you can find it there. But there is some discretion there on the part of the Ph.D. And this sample was--this sample before--it was amplified before it went through the microcon, and that amplification was also later on typed. So the sample wasn't exactly done in duplicate, but the results were obtained twice.
Dr. Cotton, in fact, two people reviewed this same reagent blank and both people concluded that there were DNA dots present, correct?
All right. And in fact, I believe you said a moment ago that you should try to do reextractions of the samples, but I believe it's your position it's not essential; is that correct?
Well, in this case, I don't know that--again, I would have to go back into the notes. I can't tell you whether or not we had enough sample available to do a reextraction.
Well, to the best of your knowledge, did you make any effort at attempting to reextract those samples?
In order to answer your--well, I know that we didn't, but what I can't tell you without going back into the record is whether or not there was anything that was available to us to go back and reextract.
Would you please take a look at your notes to see whether or not any effort was made by Cellmark to do a reextraction on those samples that were typed at the same time that the reagent blank tested positive?
I'm sorry. I guess I didn't understand when you asked me the first time. I thought you were referring to just sample item no. 7.
No. I'm asking you whether or not efforts were made to do a new extraction from item 78, 52, 56, 47 and 12 after you got a negative control failure on the reagent blank.
We didn't get a reagent blank control failure for those items because the reagent blank typed with those items prior to putting it through a microcon was not positive and those items were not repeated.
KEY QUOTEDr. Cotton, didn't you say a few minutes ago that one possible explanation for the DNA dots appearing on the reagent blank could be that when the initial extraction occurred before you even reprocessed it, that when the initial extraction occurred, that's when the contamination occurred? Didn't you say that, doctor?
Yes, I did. But the reagent blank dots don't show up. If they don't show up with a set of samples, they can't have influenced anything. If you can't see anything, it can't have affected the types.
Well, Dr. Cotton, you would agree that somehow, there was a contaminant that got into that reagent blank that created some DNA to be present; isn't that right?
Somehow, a contaminant got in there to cause the reagent place to be positive at the point for item 7 after the microcon concentration had be--had happened.
And when you did the--when you created the reagent blank, you created it at the same time, that is, you extracted those other samples I just described; namely 78, 52, 56, 47 and 12; isn't that right?
It was started at the same time, yes. But the--the end result, the reagent blank that gave the end result with two faint dots on it was not--had gone through an additional procedure that item--the other items handled apart from item 7 did not go through.
But finally, Dr. Cotton, didn't you say that you can't rule out the possibility that the contamination of the reagent blank occurred prior to that additional step, that it occurred during the initial extraction?
Do you even know to this day what actually happened to precipitate that reagent blank from lighting up with DNA dots?
No. It would be impossible to determine that at the time you did the work or at this point.
KEY QUOTENow, I ask you to turn your attention to item no. 78, the stain removed from the bottom of the boot on Mr. Goldman.
I believe you have a report on December 5th that refers to both, but I may be mistaken. Let me ask you this. By December 5th, you had completed both the RFLP testing and the PCR testing; isn't that correct?
And I believe it was your testimony on direct examination, Dr. Cotton, that the banding patterns that you observed in item 78 could be consistent with a banding pattern of Nicole Brown Simpson; is that right?
But I believe you also said that with respect to the other band seen, it was simply inconclusive as to whether or not Mr. Goldman is included or excluded; is that correct?
Now, let's turn to the fingernail scrapings. You did not actually receive fingernail scrapings from the Los Angeles Police Department, did you?
And you did not receive fingernail cuttings from the Los Angeles Police Department, did you?
What you received were solutions if you will from the Department of Justice DNA laboratory in Berkeley; is that correct?
And you do not know what portion of the scrapings were utilized to do the DNA typing, do you?
Yes. You don't know from where in the fingernail scrapings the sample was taken for you to do DNA typing in your laboratory; isn't that correct?
And you do not know from where on the fingernail cuttings the samples were removed for you to do DNA typing in your laboratory?
And would it be fair to say that you do not know whether the samples or the portions of the samples that Greg Matheson utilized for EAP typing at the LAPD laboratory could have been different portions of those samples?
One second. Your Honor, with the Court's permission, I would like to show the witness this slide which is already in evidence, 1143-E--d.
Are you aware, Dr. Cotton, that in this particular case, an EAP test was done by Gregory Matheson which provided a result that was different than the EAP profile for Nicole Brown Simpson?
Okay. And would you agree, Dr. Cotton, that when you are doing an EAP test, you are looking at what is known as a red blood cell protein?
And just to go back for a second, Dr. Cotton, to clarify it for me, in blood cells, there are two--I'm sorry. In blood, there are two types of cells, correct?
--if you want to get real fine differentiations. But if you want to classify them as red cells and white cells for our purposes, that's going to work fine.
Okay. And, Dr. Cotton, it's only the white cells that have the DNA; is that right?
I believe one of the statements you made on direct examination, Dr. Cotton, was, the more genetic markers you look at, the more information you have. Is that a fair statement?
And would you agree, Dr. Cotton, that that would apply to conventional serological markers as well as to DNA genetic markers?
For instance, Dr. Cotton, would you agree that you could include somebody with a PCR DQ-Alpha type and then, for instance, do an ABO test which excludes a person?
Yes, you could do that. I don't know enough about ABO testing to know the intricacies of what exact kinds of results you get. But could you exclude--could that phenomenon happen? Certainly.
And--well, and when you're looking at an ABO test, you're not looking at that person's ABO typing on a DNA level; isn't that correct?
And would you agree, Dr. Cotton, that likewise, the fingernail scrapings could appear to be consistent with Nicole Brown Simpson for certain DNA markers, but she could be excluded based on a red cell protein marker, that that's possible?
Let me ask you one other question about these fingernail scrapings, Dr. Cotton. You had said on direct examination that sometimes the scientists in your laboratory are asked to examine a lane from the electrophoretic plate without being influenced by the bands they see on the lane to the left or the bands they see on the lane to the right. Do you recall saying that?
And you mentioned that--I believe you said well, one of the things you want to avoid sometimes is the term called "examiner bias"?
And not just in forensic context, but have you heard that term used in conjunction with people who are asked to score autorads?
Again, it's something that I've been asked about sometimes in court. It's not something we normally are discussing in the lab. But--I don't have anything more to say about it than that.
Okay. Well, would you agree, Dr. Cotton, that one way to avoid examiner bias on a particular lane that you're asked to score is to do what you said earlier on direct examination; namely, cover up the information on the adjacent lanes and just look at the lane under question?
From what we do in the lab with autoradiographs, we do that sometimes because sometimes it's helpful in determining whether a band is sufficiently dark to score it.
Okay. Now, I want to show you an item which is People's 224-B in evidence which Mr. Matheson was asked to score using that same method; namely, where you cover up the adjacent lanes and interpret it. And I'm going to ask you to take a look at that same lane, Dr. Cotton.
Dr. Cotton, have there been times when you have been asked to look at electrophoretic plates or photographs of electrophoretic plates to determine when a certain band may be controversial as to its existence or nonexistence, to make a call on that by covering up the adjacent lanes?
Separate and apart, Dr. Cotton, for a particular type of test, would you agree, however, that it is a recognized procedure when you have a difficult to interpret lane to cover up the adjacent lanes and score it blindly?
Well, again, I can only speak for what we do in our lab, and what we do in our lab only pertains to DNA testing. So in our lab, that's a procedure that we use. I don't--I can't really speak to anything else.
Well, but Dr. Cotton, even in the field of medical research on DNA testing, isn't that a procedure that is also sometime used in that application?
Well, again, I can only tell you what I've done. Umm, I didn't do that when I was running RFLP gels when I was doing research. I just didn't. We do it in our lab for the--the purpose that I described of trying to determine whether we can all see a particular band, and that's the extent of my use of that procedure.
And would it be fair to say that you at least would regard it as an appropriate procedure for the use in your laboratory?
Now, I would like to ask you some questions about proficiency testing, okay? You had referred on direct examination to the fact that Cellmark undergoes proficiency testing; is that right?
And you referred to different types of proficiency testing that Cellmark has undergone; is that correct?
And isn't it true, Dr. Cotton, that the Cellmark laboratory makes available a summary of your laboratory's error rate as measured by proficiency testing?
We make available a summary of the lists of the number of tests, the number of samples that were involved. And if you were to calculate error rate using a specific formula, that all--that calculation is also on that summary sheet.
--specific numbers at this time, Dr. Cotton. All I'm saying is, don't you in fact on that summary that you send out actually refer to it as an error rate?
Okay. And would you agree, Dr. Cotton, that one of the purposes of proficiency testing is to provide a measure of error rate just as you did in the summary that you handed out?
You can do that with the results. In my mind, that's not the main purpose, but it's certainly a calculation that one can do.
I'm not asking you whether it's the main purpose, Dr. Cotton. I'm asking if that's one of the purposes.
And would you agree, Dr. Cotton, that there are different types of proficiency testing?
Now, Dr. Cotton, would you agree that one type of proficiency test is known as open proficiency test?
I'm going to call this chart "Proficiency Test" if that's all right with you because we're going to be describing proficiency tests. Is that okay?
And another type of proficiency test is called a blind proficiency test; is that right?
And the difference between an open test and a blind test, Dr. Cotton, is that in a blind test, the laboratory is unaware that it's being tested; isn't that correct?
And in an open test, you're aware that it's a test as opposed to actual casework?
And furthermore, Dr. Cotton, there's another way to break down proficiency tests, isn't there? Well, let me ask you this way. Other than simply dividing those tests which are open from those tests which are blind, can you also separate those tests which are internal from those tests which are external?
And in an internal proficiency test, the proficiency test is actually created by the laboratory itself that's going to be tested?
And on an external proficiency test, some outside agency is going to create this proficiency test and give it to the laboratory such as Cellmark to be tested?
And so you would agree that there are basically four types of proficiency tests based on what we've just discussed. Is that a fair statement?
Okay. All right. And a second type of proficiency test will be called an open internal test, namely, a test where the laboratory knows it's being tested and the test was made inside the laboratory; isn't that correct?
And based on what you just testified to as to the four different categories of proficiency testing, a third category would be called a blind external test, correct?
It's been brought to my attention by my colleagues that this New York lawyer also has a spelling problem and there should be an "I" in here; is that correct, Dr. Cotton?
Thank you, Dr. Cotton. So would you agree that these are the four categories of proficiency tests?
Yeah. I would like to put it over on the side if I may, your Honor, because I'm going to be referring to it during the next portion of the questioning.
All right. Now, Dr. Cotton, when you prepare tests yourself--when I say "you," I don't mean you personally, Dr. Cotton. I mean Cellmark, the Cellmark laboratory--you of course are familiar with the strengths and weaknesses of your own laboratory, aren't you?
Now, but when an external agency prepares a test, they're preparing a test that's going to go out to a number of different laboratories generally; isn't that correct?
And the external agency is not simply thinking to itself, well, what are Cellmark's strengths and weaknesses. There's not how it works, is it?
Okay. And would you agree, Dr. Cotton, that a proficiency test that mirrors actual casework would be a good basis for measuring laboratory error rate as opposed to a proficiency test that doesn't replicate actual casework?
Well, for instance, if you were interested in how well the laboratory analyzed forensic samples, you wouldn't rely on a proficiency test where you received simply, you know, a statement saying that, "We're giving you four pristine pieces of white cloth which we simply dabbed a single blood drop on and we want you to type that." Would you agree with that?
Umm, well, that's pretty much like a forensic case sample. You have standards that are on a cloth that way. So--if you're talking about, are they about the same number of samples and the same types of samples, then the proficiency tests seem to be sort of the--about the same numbers you would see--receive in an average case and about the same types--I think they try to make the same types of samples that you would receive in a case. But clearly they're making the samples. They're not like picking them up from an actual crime scene.
Well, for instance, Dr. Cotton, in this case, there are several samples which are purported to be mixed stains or stains from more than one source; is that right?
And so if you wanted to have a proficiency test where you tried to replicate actual casework conditions, it would be a good thing to have some--some mixed stains thrown in, wouldn't it?
And in this case, for instance, Dr. Cotton, there is samples which are purportedly scraped up off the sidewalk. Are you aware of that?
And would you agree, Dr. Cotton, that samples that are collected from those types of locations are frequently affected by various environmental insults?
So factors such as heat, sunlight, moisture or the substrate condition itself can affect the quality of the testing; isn't that right?
And so if you wanted to have a proficiency test that replicated casework, you would like to supply the laboratory that's being tested with some samples that also have been subjected to various environmental insults, wouldn't you?
Well, you might like to, but I think for the numbers of samples that the proficiency test folks prepare, that's probably not very practical.
But if you wanted to replicate the kinds of environmental conditions you see in actual casework, that is something that would be desirable; would it not?
I don't think it's necessary. I--I'm not even sure that it's desirable because it would be--what they would be trying to do then would be make say a hundred sample sets, all of which were identical because they want to send every lab an identical sample set, and they would be doing it in a way that would lead to those things not being identical. And part of the goal of the proficiency test is to be able to compare one laboratory to the next. So it's really critical in the design of the test that all labs get equivalent samples.
Well, although it may be difficult, Dr. Cotton, would you agree that an outside laboratory, somebody creating a proficiency test, could consistently subject the samples that are going to be submitted to some types of environmental insults? Not necessarily all, but some types.
It may be possible for them to do that, but my understanding of how the proficiency tests are put together is that--a, that's not the case, and, b, in all the guidelines for proficiency tests to be made, that--I--and I don't think that requirement to make them that much like case samples is in the guidelines.
Well, Dr. Cotton, do you believe that it's possible, for instance, that when you're creating a proficiency test that's going to go out to many laboratories, that you could take those blood swatches on cotton, for instance, and put them in a room which has a uniformed temperature so they're all subjected to heat? Could that be done?
And I think you said during your direct testimony that heat is one of the environmental insults which can actually cause the DNA to degrade.
Yes. I mean, given you're taking them into a room, you could subject them to however high you can crank a thermostat up I guess. But you could make them all uniform with a particular temperature in a room.
And you could also make them uniform with respect to a certain humidity in that room as well if you could control the amount of humidity in the room, couldn't you?
I suppose you could design a room to do that, but it--I don't know. Our air conditioning doesn't control the humidity too well. So I think you'd have to have a special design--you'd have to have a special piece of equipment to do that.
You'd have to have a special piece of equipment, that is sort of like a cold room. A cold room is just like a huge refrigerator.
And, Dr. Cotton, is there a certain kind of light that can also affect the degradation of DNA?
And could you also then subject the various control specimens that you're going to send out for DNA testing to ultraviolet light at the same time? That's something that could be done, couldn't it?
You could, but I think, again, you'd have to have a special--you'd have to design a special piece of equipment or special room in which to do that kind of thing.
And would you agree, Dr. Cotton, that--well--I'm sorry. When you receive casework samples, you do not know the answer in advance, do you?
The casework samples that come to Cellmark come from some kind of an external agency such as the Los Angeles Police Department; isn't that right?
Do you agree that external proficiency tests where you don't know you're being tested on samples are the most realistic kind of proficiency tests to measure laboratory error rate?
Well, I'm asking you about which is more realistic. You agreed a moment ago, Dr. Cotton, that it's certainly more realistic when a laboratory doesn't know it's being tested than when it knows it's being tested. Would you agree with that?
It's--it's much more--I mean, on a case, you don't know that you--wait. You don't know the answer. If you receive a proficiency test that looks like a case, then of course, that's a lot more like a case than a proficiency test that you know it's a test.
Okay. And similarly, Dr. Cotton, in an actual case, you said a moment ago the samples don't come from within Cellmark, they come from an external agency; isn't that correct?
So likewise, Dr. Cotton, in a proficiency test where the samples are coming from an external agency, that too would be a more realistic kind of proficiency test to measure laboratory error rate than one in which the samples come from within the laboratory?
It is a more realistic type of test when compared to a regular case. But I don't think that it's any better or any worse for looking at laboratory errors than an open proficiency test.
Dr. Cotton, are you familiar with that portion of the National Academy of Science report, "DNA Technology in Forensic Science" entitled "Laboratory Error Rates"?
No. It would be impossible to determine that at the time you did the work or at this point.
Yes, it states that, but I can't tell you right off the top of my head if it states that as an absolute or it should be--or that that's a possibility that should be considered.
We didn't get a reagent blank control failure for those items because the reagent blank typed with those items prior to putting it through a microcon was not positive and those items were not repeated.
Yeah. I have a spelling problem too though.