Mitochondrial DNA regions HVI and HVII population data.

Data from 1393 unrelated individuals have been compiled from eight population groups: African Americans, Africans (Sierra Leone), U.S. Caucasians, Austrians, French, Hispanics, Japanese, and Asian Americans. The majority of the mtDNA sequences were observed only once within each population group (i.e., ranging from a low of 60.3% (35/58) of the Asian American sequences to a high of 85.3% (93/109) of the French sequences). Genetic diversity ranged from 0.990 in the African sample to 0.998 in African Americans. Random match probability ranged from 2.50% in the Asian American sample to 0.52% in U.S. Caucasians. The average number of nucleotide differences between individuals in a database is greatest for the African American and African samples (14.1 and 13.1, respectively), and the least variable are the Caucasians (ranging from 7.2 to 8.4). Substitutions are the predominate polymorphism, and at least 92% of the substitutions are transitions. The most prevalent transversions are As substituted for Cs and Cs substituted for As. For most population groups these transversions occurred predominately in the HVI region; however, the African, African American, and Hispanic samples also demonstrated a large portion of their C to A and A to C transversions in the HVII region (at sites 186 and/or 189). Most insertions occur in the HVII region at sites 309.1 and 315.1, within a stretch of C's. Insertions of an additional C are common in all population groups. The sequence data were converted to SSO mtDNA types and compared with population data on Caucasians, Africans, Asians, Japanese, and Mexicans described by Stoneking et al. [M. Stoneking, D. Hedgecock, R.G. Higuchi, L. Vigilant, H.A. Erlich, Population variation of human mtDNA control region sequences detected by enzymatic amplification and sequence-specific oligonucleotide probes, Am. J. Hum. Genet. 48 (1991) 370-382] using an R x C contingency table test. Differences between major population groups (i.e., between African, Caucasian, and Asian) are quite evident, and similar ethnic population groups carried similar SSO polymorphism frequencies. There were only a few SSO types that showed significant differences between subpopulation groups. The SSO data alone can not be used to describe the population genetics with complete sequence data. However, the results of the SSO comparisons are similar to other analyses, and differences in sequence data in regions HVI and HVII are greater between major population groups than between subgroups.

Effect of reference database on frequency estimates of polymerase chain reaction (PCR)-based DNA profiles.

A variety of general, regional, ancestral and ethnic databases is available for the polymerase chain reaction (PCR)-based loci LDLR, GYPA, HBGG, D7S8, Gc, DQA1, and D1S80. Generally, we observed greater differences in frequency estimations of DNA profiles between racial groups than between ethnic or geographic subgroups. Analysis revealed few forensically significant differences within ethnic subgroups, particularly within general United States groups, and multi-locus frequency estimates typically differ by less than a factor of ten. Using a database different from the one to which a target profile belongs tends to overestimate rarity. Implementation of the general correction of homozygote frequencies for a population substructure, advised by the 1996 National Research Council report, The Evaluation of Forensic DNA Evidence, has a minimal effect on profile frequencies. Even when it is known that both the suspect and all possible perpetrators must belong to the same isolated population, the special correction for inbreeding, which was proposed by the 1996 National Research Council report for this special case, has a relatively modest effect, typically a factor of two or less for 1% inbreeding. The effect becomes more substantial (exceeding a factor of ten) for inbreeding of 3% or more in multi-locus profiles rarer than about one in a million.

Estimating minimum allele frequencies for DNA profile frequency estimates for PCR-based loci.

In order that there can be confidence that DNA profile frequency estimates will not place undue bias against a defendant, 2 methods are described for estimating minimum allele frequency bounds for PCR-based loci. One approach estimates minimum allele frequencies for VNTR and STR loci using sample size and the observed heterozygosity at a locus, while the second approach, appropriate for loci typed with allele-specific oligonucleotide probes, is based only on sample size. The use of a minimum allele frequency enables compensation for sparse sampling of infrequent alleles in population databases.

Description and analysis of allele distribution for four VNTR markers in French and French Canadian populations.

Allele distributions were determined for several VNTR loci in the general French Caucasian population, a French regional population from Brittany and in three French Canadian populations, residing either in Montréal or in one of two other regions in Québec. Allele distributions are highly polymorphic in all populations sampled. Despite a well-documented genetic founder effect in one of the populations, no disequilibrium was detected over all genotypes, within and between loci, for the data bases in this study. Moreover, there were no forensically significant differences observed between estimated frequencies of 1,964 Caucasian and non-Caucasian target DNA profiles estimated in any of the groups.

A reassessment of frequency estimates of PvuII-generated VNTR profiles in a Finnish, an Italian, and a general U.S. Caucasian database: no evidence for ethnic subgroups affecting forensic estimates.

Recently, Krane et al. addressed the effect of estimating various target DNA profile frequencies in Finnish, Italian, and general U.S. Caucasian databases. They observed that, when using an "inappropriate ethnic," or noncognate, database (e.g., a Finnish target profile frequency estimated in an Italian database, and vice versa), estimates were less common than when the target profile frequencies were estimated using their own ethnic, or cognate, database, and these estimates differed by more than one order of magnitude. These differences were attributed to the effects of subgroups. We demonstrate that the differences can be ascribed to statistical artifacts that induce large biases and correlations. Additionally, we show that the differences of estimates of any specific DNA profile frequency based on allelic size frequencies from different databases become smaller when the fixed-bin rather than a +/- 2.5% floating-bin approach is used.

Greater differences in forensic DNA profile frequencies estimated from racial groups than from ethnic subgroups.

A National Research Council (NRC) report asserted that genetic diversity between subgroups within a race is greater than between races. While this is counterintuitive to the basic tenets of human population genetics, the forensic application of statistical estimates of DNA profile frequencies has been questioned. The general approach used for estimating DNA profile frequencies by the majority of North American forensic laboratories is the use of the product rule on fixed bin allele frequencies derived from general population group databases. An informal analysis of the differences in frequency estimates within Hinf I-generated RFLP data determined in three ethnically distinct population groups--Norwegians, Spanish, and Turks--and in Caucasians and Blacks from the United States is presented. The VNTR loci analyzed are D2S44, D7S21, and D12S11. The data demonstrate that the assertion by the NRC is unfounded and that major population group databases, in lieu of subpopulation databases, can be used to provide estimates of DNA profile frequencies without consequences of wrongful bias.

Evaluation of Hinf I-generated VNTR profile frequencies determined using various ethnic databases.

Concerns have been raised about hypothetical problems arising from the use of statistics for determining the likelihood of occurrence of DNA profiles for forensic purposes. A major contention is that reference databases based on subgroups of a major population category rather than on general (or major) population groups, might yield large differences in the estimated likelihood of occurrence of DNA profiles. This hypothetical issue is based on the assertion by some people that the differences among subgroups within a race would be greater than between races (at least for forensic purposes). To evaluate the effects of the above concern the likelihood of occurrence of 615 Hinf I-generated target DNA profiles was estimated using fixed bin frequencies from various ethnic databases and the multiplication rule. Based on the data in this study, differences in allele frequencies at a particular locus do not have substantial effects on VNTR profile frequency estimates when subgroup reference databases from within a major population group are compared. In contrast, the greatest variation in statistical estimates occurs across-major population groups. Therefore, the assertion, by some critics that the differences among subgroups within a race would be greater than between races (at least for forensic purposes), is unfounded. The data in the study support that comparisons across major population groups provide valid estimates of DNA profile frequencies without forensically significant consequences. The data do not support the need for alternate procedures, such as the ceiling principle approach, for deriving statistical estimates of DNA profile frequencies.

The assessment of frequency estimates of Hae III-generated VNTR profiles in various reference databases.

The likelihood of occurrence of 1964 Hae III-generated target DNA profiles was estimated using fixed bin frequencies from various regional and ethnic databases and the multiplication rule. The databases generally were from the following major categories: Black, Caucasian, Hispanic, Oriental, and American Indian. It was found that subdivision, either by ethnic group or by U.S. geographic region, within a major population group did not substantially affect forensic estimates of the likelihood of occurrence of a DNA profile. As expected, the greatest variation in estimates for within-group estimates was among American Indian databases. Because the greatest variation in statistical estimates occurs across-major population groups, in most cases, there will be no unfair bias applying general population database estimates. Therefore, based on empirical data, there is no demonstrable need for using alternate approaches, such as the ceiling approach, to derive statistical estimates. The current practice of using general population databases and the multiplication rule provides valid estimates of the likelihood of occurrence of a DNA profile.

A comparison of the fixed bin method with the floating bin and direct count methods: effect of VNTR profile frequency estimation and reference population.

When the results of a forensic comparison of highly polymorphic variable number tandem repeat (VNTR) loci fail to exclude a suspect as a possible contributor of biological evidence, it is desirable to convey to the trier of fact the significance of the match. Furthermore, in a forensic context, it is desirable that the estimated frequency of occurrence be conservative, that is, that any uncertainty in the estimate will favor the accused. Using an empirical approach with a data base of 2046 individuals belonging to one of four population groups, this study examined the effect of the method used to estimate frequency of occurrence of a VNTR profile from a reference data base, and the consequences of using a data base that may not represent the circumstances of the crime. The fixed bin method was at least as conservative as the floating bin and genotype counting (direct counting) methods. Secondly, for forensic purposes, profile frequency estimates from different reference populations do not deviate greatly. VNTR profiles are rare in any of the data bases.

The forensic significance of various reference population databases for estimating the rarity of variable number of tandem repeat (VNTR) loci profiles.

The likelihood of occurrence of 1,964-HaeIII-generated target DNA profiles was estimated using fixed bin VNTR frequencies from various Caucasian, Black, and Hispanic databases and the product rule. The data in this study demonstrate that for forensic purposes there are smaller differences in statistical estimates of DNA profile frequencies among subgroup databases than among estimates across major population databases. This observation does not support the premise asserted by the NCR Report (1992) that the differences among subgroups within a race would be greater than between races (at least for forensic purposes). Therefore, the data do not support the need for alternative procedures, such as the ceiling principle approach (NRC Reports, 1992), for deriving statistical estimates of DNA profile frequencies. Comparisons across major population groups provide reasonable, reliable, and meaningful estimates of DNA profile frequencies without forensically significant consequences.

Fixed-bin analysis for statistical evaluation of continuous distributions of allelic data from VNTR loci, for use in forensic comparisons.

The detection of DNA polymorphisms by RFLP analysis is having a major impact on identity testing in forensic science. At present, this approach is the best effort a forensic scientist can make to exclude an individual who has been falsely associated with an evidentiary sample found at a crime scene. When an analysis fails to exclude a suspect as a potential contributor of an evidentiary sample, a means should be provided to assess suitable weight to the putative match. Most important, the statistical analysis should not place undue weight on a genetic profile derived from an unknown sample that is attributed to an accused individual. The method must allow for limitations in conventional agarose-submarine-gel electrophoresis and Southern blotting procedure, limited sample population data, possible subpopulation differences, and potential sampling error. A conservative statistical method was developed based on arbitrarily defined fixed bins. This approach permits classification of continuous allelic data, provides for a simple and portable data-base system, and is unlikely to underestimate the frequency of occurrence of a set of alleles. This will help ensure that undue weight is not placed on a sample attributed to an accused individual.

Tracking the violent criminal offender through DNA typing profiles--a national database system concept.

Implementation of standard methods for the conduct of restriction fragment length polymorphism analysis into the protocols of United States crime laboratories offers an unprecedented opportunity for the establishment of a national computer database system to enable interchange of DNA typing information. The FBI Laboratory, in concert with crime laboratory representatives, has taken the initiative in planning and implementing such a database system. The Combined DNA Index System (CODIS) will be composed of three sub-indices: a statistical database, which will contain frequencies of DNA fragment alleles in various population groups; an investigative database which will enable linkage of violent crimes through a common subject; and a convicted felon database that will serve to maintain DNA typing profiles for comparison to profiles developed from violent crimes where the suspect may be unknown.

Applying highly polymorphic variable number of tandem repeats loci genetic markers to identity testing.

The detection of alleles of variable number of tandem repeats (VNTR) loci by restriction fragment length polymorphism analysis has become an important aspect of genetic characterization for identity testing. Some VNTR loci are so polymorphic that the analysis of three to five genetic markers could potentially provide unique identity. However, the more informative a genetic marker is (i.e., high degree of polymorphisms), the better it is as an exculpatory tool. This approach currently provides the best avenue for excluding a falsely associated individual with a particular sample. When an analysis fails to exclude an individual as the source of the questioned material, a value (frequency of occurrence) should be placed on the VNTR profiles to assess weight to the inclusion in identity testing. Arbitrarily defined fixed bins were designed to accommodate quasi-continuous data and to provide a result that would not place an underestimation of the frequency of occurrence of a set of alleles attributed to an individual.

Visibility and stability of a 12-tungsten atom complex in the scanning transmission electron microscope.

A complex consisting of 12 tungsten atoms has been studied in terms of signal-to-noise (S/N) and dose response in the scanning transmission electron microscope (STEM), to evaluate its suitability for use as a approximately 1 nm resolution biological label. Molecular weight of the complex was measured as a function of radius of integration, and results were in agreement with the calculated formula weight. S/N was highest at the lowest radius of integration (0.25 nm), and decreased monotonically with increasing radius. The complex was clearly visible at a dose of 4 X 10(3) e/nm2, and exhibited negligible mass loss (approximately 8%) after an accumulated dose of 1.28 X 10(5) e/nm2. Beam-induced motion was small, 0.46 nm rms after 4 X 10(4) e/nm2. Some intensity fluctuations were observed between successive scans of the same clusters, for which a diffraction-based explanation is advanced. Upon suitable functionalization, the tungsten complex is expected to complement the undecagold cluster already in use for site-specific labeling.