MADAM CHAIRMAN ABRAHAMSON: We are going to turn in a very timely manner to Dr. James Crow who is going to give a report by the research and development working group.

Jim.

COMMISSIONER CROW: Is it better if I report from here or should I go up to the podium?

MADAM CHAIRMAN ABRAHAMSON: Wherever you're comfortable.

COMMISSIONER CROW: I'll stay here then.

Before I talk about this, I should mention a logistical question, my own problem. I will be unable to be here tomorrow and the organizers were kind enough to put me on the program today. Lisa, however, will be here. And I hope and expect fully that we'll raise questions and that will call for comment from members of the group. And if they can be transmitted to me today, wonderful, or by E-mail or some other way. But if not, and even preferably, pass them onto Lisa. And she and I, of course, are in very close touch.

The working group has met a couple of times, but I can't say that it's progress has been so great. This had better be called a non-progress report than a progress report. But I will have a few things to say, and I will proceed to do so.

The first thing to discover is that there are a number of overlaps with the legal issues group which we mutually discovered and fortunately it's been planned to have a group meeting together on December the 4th, so we'll hope to take up a number of these questions at that time.

We've had two meetings, as I've said. Most of these have been freewheeling meetings without very much resolution but we've tried to pick up the threads and I've written out a sort of tentative statement of a number of things that were discussed and some of the conclusions that perhaps will be reached. I would like to present that to you with the understanding that this is very preliminary. I will read this but I did bring along copies and any of the Commission members who want them are welcome to them.

The research and development working group met on July 20, 1998, at the O'Hare Hilton here in Chicago. Several of the members also met on October 7, 1998, in connection with a Promega Symposium held in Orlando, Florida, about which you have heard.

A large number of topics were discussed and we reached some tentative consensus on a few of them. Some items are presented here mainly to get comments from the Commission and advice as to which directions to go.

The group has not taken any formal action, so everything is tentative. We solicit advice, as I said before.

At the second meeting we took advantage of the presence of several people from abroad who were attending the Promega Symposium. These included, or were, a Dr. David Werrett, Rebecca Sparks, and Peter Gill from the Forensic Science Service of Great Britain. We also heard from Dr. Angel Carracedo from the Institute of Legal Medicine in Spain.

The British experience with handling a large felon database will provide useful information, has already, as similar problems and opportunities arise here.

The database now stars 391,000 samples -- or it did a few weeks ago -- with a backlog of roughly 100,000. The database grows at a rate of about 45,000 per year. The procedure is to put information immediately into database, but if an individual is acquitted, the sample is removed.

Most crime in the UK is committed between the ages of 14 and 21, and most first offenders are 19 or younger. Recidivism is high, but after age 30 people tend to be leaving rather than joining the criminal population.

I might add that I wish we had more detailed age distribution by recidivism or recidivism by age.

The number of reported matches that they find per week ranges from 300 to 500. We also heard in Dr. Sparks, about which I will say very little. She reported on the detailed plans for improved technology in the future, particularly with more and more use of automation. It was quite impressive to hear. I only wish I understood it well enough to describe it to you in some detail.

Although we have much to gain by the British experience, there are some important differences. One of necessity is the dealing with 50 separate states in this Country, which is not the same issue in Britain.

Our first task, our first assignment, as I understood it, is to do some technology projections for the next two, five and ten years. One of our members prepared a chart giving this kind of information or our guesses about this kind of information, and I have that available for distribution for anybody who wants it, with the clear understanding that this is the first cut as is everything else I've been saying.

This table is based mainly, but not entirely, on the discussions of the July 20 meeting with some modifications suggested by group members. This is a working document and undoubtedly will be modified with further discussion.

To summarize a few of the things, we expect that more and more laboratories will shift from RFLPs to STRs. The trend is strongly underway right now, including the use of florescent detection and multiplex systems. The trend has been underway for some time.

RFLPs, having a large number of alleles, provide much more information per locus. The limitation, however, is that the number of loci is not large and the need for binning procedures introduce troublesome statistical complications.

STRs have the advantage of almost always providing an unambiguous profile. Furthermore, they are amenable to PCR procedures, so therefore much smaller quantities of DNA can be used.

The limitation of the smaller number of alleles per locus is offset by the much larger -- in principle, almost an unlimited -- number of loci.

This is particularly true if close relatives, and particularly, particularly true if sibs are involved. I will come back to this, but I will mention it right now. That a pair of sibs always share 1/4th. That is, the conditional probability of a second sib sharing the first sib's genotype is 1/4th of any particular locus, irrespective of the rarity of the gene alleles. So that gives us a target relationship to strike at, if we are looking for how many loci is required for certain definitive level of testing.

We can expect multiplexing quickly to become the rule. It's already beginning to. CODIS has established 13 STR as the canonical loci, and we expect, as well as along with the process of obtaining allele frequencies for these loci, in a number of populations.

The definition of individual populations is being continuously defined. For example, the East Asian group is now divided into Japanese, Chinese, Korean and Vietnamese.

We can anticipate that most laboratories will soon be prepared to employ some or all of the 13 loci. The FBI or allele frequencies for those 13 loci are promised to be publicly available in about two years and we expect increasing usage for these data for calculations.

We expect STRs, and these 13 loci in particular, to dominate for at least five years and possibly longer. During this period we also expect increased automation and miniaturization. In part, this will be automation and miniaturization of procedures already in use, the same kind of STRs. In part, it will be discussion of new kinds of systems such as chip technology and SNPs, Snips as they are called in the jargon. I'll come back to this.

The CODIS database from convicted felons now has, we are told, over 190,000 RFLP profiles of convicted felons. The database for STR loci is slowly increasing with the anticipation of greater use of these.

We expect it to be well established in five years. But that depends to a large extent on funding, as some of the others groups have discussed. By that time there should be linking among the various States. For comparison, as I said, the British database now has more than 500,000 profiles.

It will probably be possible in five years to make international comparisons. Some can be made right now, because the six STR loci routinely used in the United Kingdom are included in the core of 13 FBI loci. So some comparisons are possible now.

In five years the human genome project will be essentially complete. This will generate a host of new possible markers. By ten years, new systems will certainly be possible.

A serious issue, and one which we must surely address, is the extent to which laboratory systems should keep up with the state of the knowledge. The CODIS markers systems needs to have some stability, because we can't go around changing the system every five years when the time and money investment is so large. In ten years the new systems superimposed on the existing ones will provide new opportunities and then they may also lead to confusion.

We expect greater automation of laboratory procedures and computerized analysis. This should increase the accuracy and reduce the expense.

I say that without full confidence. It seems to me the technical innovations which are supposed to reduce the expense hardly ever do. But we can still hope.

This should increase accuracy and reduce expense, as I said. It will not, of course, replace human judgement. One place in which computer analysis is likely to play a larger role is with mixed samples for which the calculations can be complicated.

There seems to be little need for a recommendation from our group on developing new systems and improving existing ones. New techniques are being aggressively researched by private laboratories, better laboratories, and universities, quite regardless of what we may well say.

The issues are how rapidly these better methods, which are sure to come, should displace an expensive system which is already in place. Databases based on STRs should be around for a long enough time to offset the cost of setting them up.

A note to myself to remind you people, we were wondering how much technical details should go into a report like this. And one possible disposition of this is to have an appendix in which the technical procedures now and anticipated for the foreseeable future are written in appendix form so as not to interfere with the telling of the story, but nonetheless to be available for those who would like it. We have plenty of expertise on the Committee, whether it's writing skill or not I don't know, who can write up these report as an appendix. Let me have advice from you, not necessarily now, as to how much this should be a textbook in other words.

Two other aspects I want to talk about technically. Mitochondrial DNA markers have been available for some time. These have two advantages. One is that since there are many, many mitochondria per cell, samples that are too small or too degrading for nuclear analysis may be informative for mitochondria. So we can expect, and we already are finding, increasing use.

The second advantage is that mitochondria are inherited through the female line and this often permits tracking ancestry and remote relatives. This was used, as all of you in the room know, to identify the Romanoffs a year or so ago.

Until recently there was no way to track the male line of decent. This is no longer true I'm happy to say. There are now several good Y chromosome markers and we can expect more.

Mitochondria and Y chromosomes are one area where SNP technology can be especially useful. The Y chromosome is turning out not to be the depauperate blob of DNA that it was once thought. It determines maleness, but we didn't know anything else for a long time. And there are not any good genes on the Y chromosome, but there are a lot of good molecular markers that are turning up day-by-day.

These gender-specific methods may well find their major use outside the criminal field, as exemplified by the recent study of Thomas Jefferson's descendents.

A little bit about improvements and reliability, the robustness and speed. The working group has had little discussion of this subject. One reason for not doing a lot with it is the existence of the DNA advisory board and the exact jurisdiction between these two is an interesting question for some thought.

We need to consider, in any case, improvements in technology and protocols and in training. We expect in the five year view to have greater automation with resulting increase in accuracy. Over the longer period there will be technology for sample handling and analysis at the scene. A prompt determination of the profile at the crime scene can speed the identification of the suspect, and perhaps more important, quickly eliminate innocent persons with important saving of time and human anguish.

Despite improvements, technical and human error are likely to remain considerably more probably than random matches. As techniques become ever more sensitive, small amounts of stray DNA, possibly planted, can become an increasing problem, and we hope to address that.

The most likely human error that could lead to false incrimination is a sample switch. This can be reduced, as can all other errors, by scrupulous attention to details. Also in performing two PCR reactions with a couple of redundant loci between the two, will permit detection of some possible mix-ups.

But finally, as was emphasized in both NRC reports, every effort should be made to preserve material for subsequent re-testing should there be the need for this.

We want to say something about the question of inferring ethnicity or phenotypic characteristics from a DNA sample. It is already possible in many cases to state the race of a person contributing a DNA sample with a high probability of being correct. With the canonical 13 STR loci, this should be practical in a great many cases I think right now.

I took one example. I chose it, so it isn't really random. But I picked three STR loci. And with those three loci, and particular alleles in them, the frequency of this profile in the black population was about three per million, and in a white population about two in ten billion.

The likelihood ratio here is 2,500. So the probability is 2,500 times as great of finding this particular type in the white population as in the black population. That must be considerably more accurate than eye witness identification of race that is ordinarily used as a way of finding leads.

I notice, however, on reading the report of the legal issues group that they reported a likelihood ratio of 2,500, that gave the wrong answer. Personally, I'm curious to see the data. I don't think there have been enough possible tests of this particular procedure to make a mistake of 1 in 2,500. But in any case, one doesn't expect it to be right.

But the use of purely statistical manipulations with existing data is a possibility right now and as a way of finding leads it offers considerable promise. It also raises ethical and social issues as will clearly come up.

We need to, and I hope we can, do the kind of proper calculations here. Really try all possibility combinations, or at least a very large sample of possible combinations, to see how often one can expect a significant likelihood ratio between two different racial or ethnic groups. The larger the number of loci, of course, the more reliable this particular procedure is.

The question of phenotypic characters is more problematic, both ethically and genetically. The loci now used for forensic analysis are likely to be now used and likely to be used in the future are not associated with any external traits.

And it's an important consideration in developing new systems that what's used as DNA markers for this kind of investigation not be associated with known genes that had any phenotypic effect.

So determining that, for example, a DNA sample is left by a person with red hair is not practical now. How soon it will be, I don't know. The number of common traits that are caused by single gene differences is not likely to be high. It's certainly not high and I think it's not promising that it will ever be high. There may be rare traits, however, that are identifiable.

Using polygenic systems or identify traits based on polygenic systems seems less probable, but I hope we can discuss this. The genome project and other research can be expected eventually to identify some genes leading to identifiable traits.

But this involves a quite different kind of research than goes on for the detecting of new forensic systems, because in the detection of forensic systems one deliberately looks for traits that are not associated. I mean, looks for markers that are not associated with traits. And I, for one at least, hope that that tradition can be maintained.

I will mention at this time again that this is one of the issues that we certainly want to discuss with the legal issues working group, and we have a meeting schedule for December 4 to raise this and other questions. I hope you will ask members of either group for questions you would like to have discussed.

The last part I want to say a bit about has to do with population structure and the existence of near relatives in the database. The question of relatives did come up in our discussion.

There is always the possibility of a suspect being a relative of the source of the evidence sample. Such relatives may not know of each other's existence, particularly if they are half sibs. There are probably quite a number of people who are half sibs and don't know it.

For more remote relatives this is not much of an issue, but for close relatives, it is. I have done some calculations based on four loci of the STRs and just how often, I ask how often one would find a match of unrelated persons and different degrees of relatives, and let me give just a few sample numbers. I did this for African Americans, Caucasian, and Hispanic populations.

Incidentally, the document from which I'm reading here, I have extra copies and I'm certainly happy to have the Commission members have them.

For example, I took the four loci that I had that were easily available and calculated it. And then I extrapolated to 13 loci, on the assumption that these four were representative. Actually, the four are a little more sensitive than the rest of the group, so perhaps these are slightly off from that magnitude, slightly less sensitive.

With 13 loci, the average probability of a random match is 2 to the minus 13th power. That's the reciprocal of 2 followed by 13 zeros. It's as small a number as we're likely to want to know.

How much difference does it make if the population is structured. We have used, largely following Bruce Weir, the symbol theta to stand for the measurement of population structure. This is essentially the same as Sue O'Rides (phonetic) quantitivity called AFST.

The estimates for these values are usually less than 1 percent. And if I calculate from the equations in our 1996 report, these make no difference of the size of an order of magnitude, and especially if we use the conditional probability equations of Balding, Balding and Donnelly. That changes this probability from about 2 to the minus 13th to about 4, for a theta value of .01.

So it looks to me as if most of the time doing these corrections will not change the calculation by an order of magnitude although if it turns out we have underestimated the degree of population subdivision, then it might change it by an order of magnitude, or even two.

What about half sibs and what about full sibs. Half sibs, that drops from 2 followed by 13 zeros to 2 followed by 10 zeros. So it's still very unlikely that half sibs would share a phenotype profile.

What about full sibs? Let me go back to full sibs for just a minute. It's possibly obviously, possibly not, that full sibs are different from other relatives because two persons who share both parents can have two sets of genes and share these two sets of genes. That's not possible with other degrees of relationship unless the gene happens to be common. So this means that for full sibs, irrespective of the rarity of the gene, there is a conditional match probably of 1/4th. It would be slightly larger than 1/4th because other genes contribute, but the 1/4th is the major factor in this.

In any case, that calculation, the probability of full sibs matching the average probability and the STR database is about 1 in 10,000. If you raise that to 15 loci instead of 13, that drops to about 1 in a million.

So what I'm interested in concluding from this rather cursory analysis, fortunately we have some data from Jack Rubarti (phonetic) who loves to analyze large bodies of data, in which the kind of things that I've done on a small sample can be done on a large sample or even the extended set of data.

Two points stand out. One is that the theta correction makes very little difference if one uses the value of .01. We already knew that, but I'd like to point it out again. Second, with additional loci even full sibs have a small match probability.

Paradoxically, as the data on various sub-populations become increasingly available, increasing numbers of loci will render them less needed. An object of the much discussed interim ceiling principle of NRC1, which was so roundly condemned, was to have a procedure which was independent of racial and ethnic classification.

In looking to the future, it's not very far in the future that with an increasing number of loci this objective may be obtained in a scientifically more acceptable way. And if we have a large enough set of loci that they would distinguish full sibs, of course they will distinguish individuals of any lesser relationship and it will be no longer necessary to ask what ethnic or racial group this particular database comes from.

One of our tasks is to ask how soon this utopian state will be arrived at. Thank you, Shirley. That's all I have to say right now.

MADAM CHAIRMAN ABRAHAMSON: Thank you, Jim.

Are there any questions or comments or suggestions?

Barry?

COMMISSIONER SCHECK: What was the view about mitochondrial testing in terms of its discriminatory power? Did you look into that at all?

COMMISSIONER CROW: The discriminatory power is not large compared to a large number of nuclear loci. It's large compared to any one of these. Maybe that's all I can say is it's quantitative. I'll tell you what you already know, but the usefulness of mitochondrial DNA I don't think is so much for forensic kind of purposes as it is for the various other kinds of uses that are certain to be found.

COMMISSIONER SCHECK: I don't think that's true at all. In fact, what we're seeing increasingly is so many of these cases can turn on the mitochondrial analysis of a single hair.

And what's becoming interesting to try to do -- and we've run into this problem in a number of our innocence project cases -- is that sometimes one technological challenge -- and I throw this out to you because I've gotten a lot of different answers on it -- is the extent to which you might be able to extract mitochondrial DNA from a semen stain, particularly a mixed stain, so you can get the male pattern and see if you can match that to the mitochondrial DNA of a hair.

I see a lot of cases where that becomes critical in order to make a case. And I think what you're finding forensically -- I mean, I really commend your group to look at this -- is that many investigations are turning on the analysis of extracting mitochondrial DNA from a hair.

COMMISSIONER CROW: I realize, and I guess I said it, that mitochondria does have the tremendous advantage that there are many more mitochondrial particles per cell, so that a tiny amount of tissue can yield a result. But it's still not as powerful.

COMMISSIONER SCHECK: It's not as powerful, but what I'm saying is forensically it's becoming extremely important, particularly if you have rapists, for example, that use condoms. Which, in the year of DNA, many law enforcement people are warning us about, we're going to be making cases based on hairs.

COMMISSIONER CROW: Pure sperm shouldn't have much mitochondria, but there are enough other tissues around to have it. I should think the Y chromosome is going to be increasingly useful in this context.

COMMISSIONER BASHINSKI: You had a question about the level of technical detail that should be in this. Listening to your description of where we are with STRs and where we're going, I think it's going to be very important for this document to clarify for the general public and for the legal community the degree of acceptance of these -- that this technology has and the widespread use that is coming with technology. And I think we do need to have clear explanations, at least in an appendix, of how these technologies are different and how they are similar. Because I think there is a lot of misunderstanding and a lot of confusion whenever we start talking about new technologies. Sometimes it isn't really that much newer, and we don't want to confuse people.

Anyway, I think the whole issue of what is going on now is going to be very important to make very clear as a follow-up to NRC's work.

COMMISSIONER CROW: I rather expected to hear that response from a large number and I think we probably ought to do it.

One dilemma that we have -- everybody does -- but in writing the NRC2 report we were constantly admonished to make the report simple so any child can understand it. And then as soon as the critics landed on this, they accused us, quite properly, of over-simplification, and therefore making some mistakes. They are minor mistakes, but they're mistakes. So there is a fine line to follow here.

MADAM CHAIRMAN ABRAHAMSON: So you use appendices.

COMMISSIONER CROW: I think appendices is a good way to do it.

COMMISSIONER THOMA: I was going to wait to talk with you later, but I agree with Jan that that would be very appropriate.

COMMISSIONER GAINER: I was simply going to affirm that. And I think one of the audiences of that are the CEOs of either police departments or potential lab directors and members of the general assemblies when they try to figure out how much money ought to be put in this and when.

So it's especially important when you talk about 5 years, 10 years, 15 years. Because there is a misconception if you invest today, they can't see however you'd want to be changing your strategy and changing the number of forensic scientists you need.

And that has to be done very simply so the members of our elected officials can understand that.

COMMISSIONER CROW: Well, another point that I probably should have said, and that is I think we should clearly list the advantages and disadvantages, because every system has both.

MADAM CHAIRMAN ABRAHAMSON: Were there any other comments?

COMMISSIONER THOMA: No. That's the only comment I really have.

DIRECTOR ASPLEN: I would just say for any of the Commissioners who are interested in that meeting on December 4, as Dr. Crow said, it is an important issue. The interconnection between the work of the legal issues group and the research and development group. That meeting is going to be here in Chicago, but it will be at the airport. It will be at the Hilton.

No, it's not at the Hilton. But it will be near the airport. And as such, in an attempt to make it more convenient for folks to fly in or fly out.

DR. FORMAN: It will be at the Comfort Inn, which is a stone's throw from the airport. A new facility we are told.

DIRECTOR ASPLEN: But what I wanted to do is make the offer, as we always do, to all the Commissioners who would like to attend that meeting. All you need to do is to let us know. Call Robin and we will make sure that the travel contractor gets in touch with you so that any of the Commissioners who want to go to that can in fact do that on December 4. I believe that's on a Friday. Again we made it close to the airport so that you can fly in and fly out the same day and reduce the travel time.

COMMISSIONER CROW: One of the speakers that we had at our Promega meeting was Peter Gill from the forensic service. He didn't tell us a thing about what they are actually doing, but he is a member of the Commission in Britain to look at this question of identifying traits as opposed to just racial. So I think we have to look at it.

COMMISSIONER SCHECK: Red hair, I was surprised because when I was over in the UK they said they made a lot of progress on red hair. Dr. Weir said that to us as well when we were last time in Chicago.

COMMISSIONER CROW: I think any human geneticist, even me, would say red hair is the most promising single trait to be monogenic and conspicuous. I think eye color, skin color doesn't mean much. So it's hard. It's hard to think of good -- albinism perhaps could be used, but that is a fairly rare trait.

MADAM CHAIRMAN ABRAHAMSON: Is there any other comments, suggestions?

Well, thank you very much. And not only is everyone invited to the December 4 meeting, but the staff will circulate a calendar for the year to everyone showing all the meetings as of that date, all the working group meetings as of the date. And we will keep updating that as we go along so that everyone can know that.

Thank you very much Jim. And Jim -- I'm sorry, Barry?

COMMISSIONER SCHECK: I just had one thought. One of the things that when I was in the United Kingdom looking at their databases and their figure of 3 to 500 hits per week that struck me, was that that number, by their own admission, is slightly inflated. Because when people are arrested in the UK, they don't have the same fingerprint capacities that we do to identify people who are using aliases. So some of the matches are matching people to each other who are using aliases. And being from New York, even when I started as a public defender in 1975, we always had the capacity to use the fingerprint system to identify everyone arrested in the State with aliases.

But lately I've been traveling other places in the country and I've ascertained that that is not as uniform as I thought it was. I was in Mississippi last week, for example, and they were telling me they don't have that capacity.

What I was wondering about is to the extent to which your group can sort of give us a report comparing AFIS, the AFIS system and potentially the DNA databank system, how those things could be used jointly.

The British model is based completely on the principle of treating the DNA just like fingerprints, both legally, ethically, and practically. And they are better with DNA than they are with fingerprints it turns out.

COMMISSIONER CROW: We have reams of data from David Werrett, as you probably guess. I expect everyone has the same thing.

COMMISSIONER GAINER: What is their turn around on identification of DNA?

COMMISSIONER SCHECK: Well, they make deals. They're a very unusual institution as he explained to us the last time. They actually are -- the forensic science services makes contracts with local police departments as to how much of the services they want. And in certain areas they will contract to take a certain number cases on a high priority basis and they turn them around within two weeks.

And the reason that's important is that they found when they don't turn them around in two weeks then police find the typing less useful so the point of catching these people in some other way in many of these instances.

But they have the capacity in many of these cases to turn them around in two weeks and they guarantee it.

MADAM CHAIRMAN ABRAHAMSON: But not all cases?

COMMISSIONER SCHECK: Plenty of cases.

COMMISSIONER GAINER: Unless I'm missing something, given our goal in an AFIS identification, a fingerprint is 30 to 60 minutes. Is that what you're shooting for now, Superintendent?

COMMISSIONER HILLARD: Definitely.

COMMISSIONER GAINER: It would be interesting to see how we would ever think in our system that would match up.

COMMISSIONER SCHECK: I guess what I'm thinking about more is to what extent does the AFIS system include a databank of unsolved crimes and fingerprints taken from unsolved crimes. I was very struck by the San Antonio experience. And I've asked this question a lot. I haven't seen -- I mean, how much of a cold case file is there for fingerprints in most jurisdictions? My impression is not very much.

COMMISSIONER BASHINSKI: It's extensive in California and I think most of the States that have had ongoing AFIS systems for a long time, and that's one of the primary goals is to keep your unsolved cases and then you do get hits.

COMMISSIONER SCHECK: That's what I was thinking of is finding out if we can get more data on that and see how the two databases could relate. I think that might be useful.

MADAM CHAIRMAN ABRAHAMSON: Superintendent Hillard?

SUPERINTENDENT HILLARD: Madam Chairman, I wanted to ask Dr. Crow in England, the national database, who maintains it over there.

COMMISSIONER CROW: I guess it's the forensic science service.

Barry knows more about that than I do.

COMMISSIONER SCHECK: They maintain it centrally.

SUPERINTENDENT HILLARD: When you say centrally, what agency?

COMMISSIONER SCHECK: Forensic Science Service will maintain both the printouts and the samples.

COMMISSIONER CROW: And then they store the samples, too. They have the liquid nitrogen or other refrigerated systems.

One thing that impressed me -- we'll see how much general agreement there is on this -- was how much simpler it is to do this in England than in the United States, mainly because there is just a single jurisdiction.

COMMISSIONER SCHECK: Well, you haven't been to Scotland or Ireland yet.

COMMISSIONER CROW: I forgot about the Irish.

COMMISSIONER SCHECK: They have nothing in there from Scotland just about. It doesn't seem so simple.

COMMISSIONER CROW: England itself is pretty much homogenous in this respect.

MADAM CHAIRMAN ABRAHAMSON: Because Dr. Crow will not be with us tomorrow and our public comments is not until tomorrow afternoon, nevertheless, Jim, if you would take any comments or questions or suggestions from the audience, we would appreciate it. Is that all right?

COMMISSIONER CROW: Sure.

MADAM CHAIRMAN ABRAHAMSON: Dr. Forman?

DR. FORMAN: I just wanted to clarify a little point about who maintains the samples in England. They are maintained by the Forensic Science Service, but it's as a contract with the individual police departments. So the samples are actually owned by the individual police departments and they are contracted out to the forensic science services for maintenance and caretaking. If that makes a difference.

MADAM CHAIRMAN ABRAHAMSON: Thank you.

There was someone else. Would you state your name, please.

DR. CROUSE: My name is Cecelia Crouse and I'm from West Palm Beach, Florida, and I have three comments.

One is as a red head I want to let you know I didn't do it.

But I also wanted to make you aware of what is going on in our State with regard to admissability hearings and population databases. The last couple ones that we've had we have experts down there who are saying that databases go stale after five years. And they're saying especially in States like Florida and California where there is a lot of migration in and out that that is one of the major reasons for genetic drift. And I just thought I wanted to make you aware of that, that the judges like that. They think that makes sense.

The other thing that I'm interested in is when you're considering match probabilities, do you take into account the higher mutation rates in these loci versus RFLP?

COMMISSIONER CROW: The higher mutation rates?

DR. CROUSE: With regard to siblings.

COMMISSIONER CROW: I think for pure matches it doesn't matter. If we were doing paternity testing or something like that mutation rates would matter. But I think for matches, the likelihood of mutation making up the major portion of the cells is so remote as to be ignored.

DR. CROUSE: In my family of 35 members we have three mutations.

COMMISSIONER CROW: Mutations do occur, it's just they don't confuse the analysis.

DR. CROUSE: I mean strictly with STRs. Thank you.

MADAM CHAIRMAN ABRAHAMSON: Thank you.

Are there any other comments, questions, or suggestions for Dr. Crow in that group?

Hearing none, then we will complete that and thank you, Jim.

And I'm happy to also report that one of the things that has been bothering the staff and me is that we didn't have a vice chair so that if I couldn't make a meeting or needed assistance in the interim that there was nobody. So Jim Crow has graciously agreed to be vice chair as long as I agreed to not impose any duties on him. That was our agreement.

COMMISSIONER CROW: That last clause is very important.

MADAM CHAIRMAN ABRAHAMSON: I know you began to pale visibly, so I was very careful to add that. Thank you, Jim. And you're off to your other meeting or staying for this one?

COMMISSIONER CROW: I'm going to stay for the rest of today.

Previous	Contents	Next
Back to National Commission Main Page