Introduction to Forensic Sciences 2nd Edition phần 8 pdf

Restriction Fragment Length Polymorphism RFLPs of Variable Number of Tandem Repeats VNTRs Are the Basis for the Original DNA Profiling At many locations throughout the human genome, shor

which are degraded or in some way made smaller by actions of physical orenvironmental agents or age Because theoretically it is possible to multiplysingle strands of DNA to essentially millions of copies of that single sequence,PCR is extremely sensitive, and from 1 to 5 ng of DNA can be successfullytyped using the process.23 Figure 11.10 and Table 11.6 compare the differencesbetween conventional RFLP/DNA analysis and PCR/DNA analysis.

Figure 11.9 The polymerase chain reaction: Multiple copies of specific DNA segments

or genes were originally produced by cloning, cutting out the segment wanted and inserting

it into a host cell which would, as it reproduced, make copies of the inserted gene along with its own DNA Now DNA may be copied enzymatically using a temperature-insensi- tive DNA copying enzyme or polymerase Starting with double-stranded DNA, the specific site to be amplified is targeted by using primers which flank the target site and act as anchors for the synthesis of a new DNA strand (A) The first step of a cycle involves melting the DNA to expose the nucleotides of each strand allowing the primers to bind (B) Next the temperature is lowered and the polymerase enzyme facilitates the synthesis of two new DNA strands using the old strands as templates (C) These two double-stranded DNA molecules are melted or denatured in cycle 2 to begin the process anew It is called a chain reaction, because at each cycle the number of previously existing DNA molecules is doubled.

C

A, B

©1997 CRC Press LLC

Restriction Fragment Length Polymorphism (RFLPs)

of Variable Number of Tandem Repeats (VNTRs) Are the Basis for the Original DNA Profiling

At many locations throughout the human genome, short and long sequences(two to over several hundred base pairs) of DNA bases are repeated over and

Figure 11.10 Comparison of RFLP (Southern-blot) and PCR (amplified) DNA tion tests: The power of forensic tests depends on their ability to include or exclude individuals as contributors of evidentiary samples, but the application of particular tech- niques depends on the number of tests required to produce odds of exclusion which suggest that the match means that the evidentiary sample came from the identified individual Selection of a test or set of tests also depends on the condition of evidence Fewer RFLP tests than PCR tests are needed to produce a given probability, but larger sample sizes are required for RFLP tests In the future, PCR-based sequence-like or sequence-based tests may eliminate this difference.

identifica-Table 11.6 Comparison of Two Analysis Methods for Forensically Applicable DNA Loci

8–22 alleles 250–1,000 alleles

over again.15,24 They are said to repeat in tandem, meaning they repeat tinually in a chain-like sequence The number of these repeating units ishighly variable, so that most people have differing amounts of these repeatunits inherited from their mother and father (see Figure 11.11) When thelength of the repeat units inherited from each parent will be different, thiscondition is known as heterozygosity The existence of these repetitivesequences is very important, as they are being used to locate certain genes

con-in our chromosomal structure, and also some have been lcon-inked to geneticdiseases such as fragile X syndrome and myotonic dystrophy.25 They are alsoexcellent sites to use in assessing differences among humans.26 By use of atechnique such as RFLP the forensic scientist is able to quantitate the lengths

of these VNTRs and use them to identify with high precision and accuracythe identity of an individual Because there are a limited number of thesevariations at any one site, different individuals can share similar size VNTRs

It then is not possible to determine the source of a sample from use of oneVNTR locus (spot on a chromosome) If, however, many different VNTRsare used, a profile of an individual is created with a very high discriminationpower (see Figure 11.12).27, 28 There are several ways to determine the length

of the VNTR

One is by using a restriction enzyme previously described to “cut” theVNTR out of the piece of DNA and then to separate the pieces using elec-trophoresis Another way is the use of the PCR process to copy or amplifythe VNTR pieces of less than 3000 bp and then separate them by electro-phoresis.29 This does not involve the use of restriction enzymes but the use

of PCR primers

Figure 11.11 Example of a VNTR (variable number of tandem repeat): Genomic DNA, all the DNA in each cell, contains many sections which are repetitive These often vary from individual to individual, because the number of repeats is different Because of this, the length of the repeat DNA will vary when this section of DNA is cut out and visualized

or amplified and visualized The “repeat units” can be as small as a single base pair to many hundreds of base pairs making up a repetitive sequence The sequence may be many thousands or millions of “repeat units” This kind of DNA has been referred to as “junk” DNA in the past, but new evidence may show that it may not be junk after all Thus, the sequence made up of “repeat units” from our father may differ from the sequence from our mother When there is a difference between genetically defined multiple forms of a particular character, it is called polymorphism.

©1997 CRC Press LLC

Other Types of VNTRs and Sequence Polymorphic Areas

of Nuclear and Mitochondrial DNA are Also Used

to Profile DNA in Forensic Cases

Small, amplified DNA segments are currently used in sequence-type analysis(mtDNA and MVR), fragment size analysis (STR, AMPFLP), or dot-blotanalysis (DQalpha and amplitype PM™).7,8 In sequence-type analysis, theamplified DNA fragments may be sequenced directly after amplificationusing any of the sequencing methods, or the amplified fragments may beseparated to produce a ladder which resembles a sequencing ladder Direct

Figure 11.12 Four DNA probe composites of autorads containing samples from suspect, victim, and evidence with a final frequency of occurrence of 1 in 3,400,000,000: The four autoradiograms or autorads in this figure depict matches between the suspect’s DNA patterns and those produced by the DNA extracted from the evidence Since these four pattern matches are first made visually and subsequently by computer sizing, the examiner has estimated the odds of someone other than the suspect also producing a matching DNA pattern for each of the four probes These are the numbers under each of the four autorad diagrams: 1 in 200, 1 in 303, 1 in 125, and 1 in 450 For the first match, using probe #1, the odds of finding another unrelated person with the same DNA pattern as the evidence are 1 in 200; this means that 1 / 2 of 1% of all humans will produce this pattern There are, however, approximately 5 1 / 2 billion people on this planet; so, we must consider the fact that these odds of 1 in 200 mean that approximately 27 million unrelated people will also produce DNA patterns which match the evidence DNA pattern matching power depends

on the use of three to five different probes to identify independent DNA segments The independence has been statistically verified and the examiner is, therefore, able to combine the odds for each probe to estimate the odds of finding an unrelated individual, other than the suspect, who would also produce matches on all four autorads This is done by multi- plying 200 × 303 × 125 × 450 to produce composite odds of approximately 1 in 3.4 billion These odds suggest that slightly more than 1.6 people on the planet who are unrelated to the person who produced the evidentiary pattern will match on all four of the autorads.

sequencing is most frequently done with the highly polymorphic controlregion or D-loop of mitochondrial DNA, and approximately 400 bp aresequenced.9 This is a valuable technique for identification, because all mater-nal relatives are expected to have identical mtDNA; however, it is, for the samereason, less discriminating because maternal relatives are indistinguishable.30

Minisatellite Variant Repeats (MVR)

MVR (minisatellite variable repeat sequencing) uses the fact that some VNTRrepeats have internal polymorphisms which may be used as terminators inmuch the same way that Sanger sequencing uses dideoxynucleotides to ter-minate polynucleotides in synthesis These are then separated on a gel, andthe sequence is read directly off the gel The minisatellites MS31 and MS32 arecurrently the only VNTRs being used for identification using MVR methods.31

AMPFLP (Amplified Fragment Length Polymorphism) Represents Another Type of VNTR that Is Smaller than the VNTRs Used for RFLP

In AmpFLP analysis (amflip, or amplified fragment length polymorphism),sample DNA is amplified using primers which flank a core repeat of approx-imately 10 to 20 bp (see Figure 11.13).7 The fragment length polymorphism,like VNTR polymorphism, is based on the number of core repeats found in

a particular allele Here alleles are defined as DNA segments on homologouschromosomes with different numbers of repeats For example, the AmpFLPinherited from the mother may have 74 repeats, while that inherited fromthe father may have 38 repeats AmpFLP alleles are, however, smaller thanVNTR alleles They range in size from 100 to 1000 bp, while VNTR allelesrange from 200 to more than 20,000 bp Most VNTR alleles are too long atthis time to be amplified by PCR, and the ends or termini of VNTR allelesare defined by restriction enzyme cut sites which flank the tandemly repeatedcore sequences Long PCR may change this limitation Long PCR may beable to amplify more than 20,000 base pairs of DNA in the very near future.AmpFLP loci are selected for analysis because the size range of core repeats

is efficiently amplified by PCR; consequently, the termini of these AmpFLPalleles are defined by primer sites, not by restriction sites These differences

of size explain, in large part, the fact that VNTR loci have an enormousnumber of alleles continuously distributed over the size range, while AmpFLPloci have an approximately discrete number of alleles, usually in the range

of 5 to 25

©1997 CRC Press LLC

Following amplification, the DNA sample is separated on a gel, usuallypolyacrylamide, and the amplified fragments are visualized with silver stain

or fluorescent dyes As in VNTR analysis, size ladders are included on theanalytical gels In AmpFLP analysis, however, alleles are treated as discreteunits which allows visual comparison of alleles with the ladder alleles, unlike

Figure 11.13 Examples of various repeat polymorphisms in the human genome: Genomic DNA, all the DNA in each cell, contains many sections which are repetitive These often vary from individual to individual, because the number of repeats is different Since the number of repeats is different, the length of the repeat DNA will vary when this section of DNA is cut out and visualized or amplified and visualized (A) VNTRs have repeat lengths or cores which range from 9 to 40 bp, depending on which specific gene is being examined These VNTRs are isolated and visualized for pattern comparisons by restriction enzyme digestion and DNA probing VNTRs are the genes first used by forensic labs to produce “DNA fingerprints” by combining patterns over four to six genes Although VNTRs are highly polymorphic, this testing methodology is expensive and time consuming Two additional repeat length polymorphism tests have been introduced to reduce costs and testing time Rather than cutting and probing, these tests amplify specific repeats, and the amplified sequences are visualized without probing (B) AMPFLP (amplified fragment length polymorphisms) have cores which range from 8 to 16 bp (C) STR (short tandem repeat) cores range from 4 to 6 bp.

Short Tandem Repeats (STR) Represent a Very Small VNTR

Another type of amplified repeat analysis is STR, or short tandem repeatanalysis It is very similar to AmpFLP analysis, but the repeat sequences areshorter still (4 to 6 bp) Additionally, a number of STR loci may be amplifiedand separated simultaneously, a technique known as multiplexing Thisincreases the discriminatory power of STR analysis, while decreasing the workand time involved in the analysis There are approximately 4.0 × 108 STR locidispersed throughout the human genome STRs consist of small numbers ofrepeat units, usually three, four, or five repeats, which are from 50 to severalhundred base pairs in length

Much effort is being made at present to shorten the time of analysis ofSTRs and AmpFLPs, as well as to increase the number of STR and AmpFLPtypes analyzed Several types can be analyzed at the same time and on thesame gel with sophisticated hardware and novel tags attached to the DNA.One of these attempts has been the use of fluorescent-tagged STRs andAmpFLPs Dr Ron Fourney and Dr C.J Fregeau of the Royal CanadianMounted Police have been instrumental in developing this technology (see

Probes for the DQalpha alleles are fixed to membranes as dots Thesample DNA is amplified, and allelic identification is achieved by allowingthe amplified sample to bind to the appropriate dots on the membrane Thebound sample is then visualized using a conjugated enzyme and dye Then,the allele is identified as having bound to its complement on the membrane(see Figure 11.15)

The second discrete allele system is amplitype PM™ It consists of certainprobes bound to a strip as in the case of HLA DQalpha The loci or typesare as follows: HLA DQ-alpha (separate strip), LDL receptor (LDLR), glyco-phorin A (GyPA), hemoglobin G-gamma globin (HBGG), D7S8, and groupspecific component (GC) The advantage of this typing system is that one

©1997 CRC Press LLC

Figure 11.14 The evolution of DNA typing in North America: This montage represents the evolution of DNA typing in North America from the (a) initial single locus DNA typing profile (D1S7) followed by PCR based methods including (b) AMP-Flaps (D1S80), (c) mini- satellite variant repeats ([MVRs] D1S8) and (d) fluorescent-tagged short tandem repeats (STRs, bottom: HumCD4, yellow; HumFABP, blue; HumACTBP2, green; ABI Genescan™

2500 marker, red) Contributed by the Biology Research and Development Support Unit and Richard Musgrave of the Forensic Photo Unit, Royal Canadian Mounted Police (Figure courtesy of Eaton Publishing Company.) See color plate

works with five loci or types instead of one type as with DQalpha, and thepower of discrimination is increased The average frequency of occurrencecan be as low as 1 in 400 or as high as 1 in several million See Table 11.7

and Table 11.8 for frequency data for HLA-DQalpha and amplitype PM™markers

Mitochondrial DNA (mtDNA) Analysis: The New Frontier

in DNA Profiling and the Sequencing of Polymorphic Sites

of the Nuclear and Mitochondrial Genome

Mitochondrial DNA is located outside the nucleus of the cell in the producing mitochondria The advantage of this type of DNA is the greatnumber of mitochondria per cell A single hair root has been successfullytyped using mitochondrial analysis The information in DNA is encoded inthe linear array of nucleotides The genetic code consists of nucleotide tripletswhich may be converted in transcription and translation to amino acids toform proteins Ultimately, then, it is reasonable to expect that DNA sequenceinformation will be used by forensic scientists

energy-Currently the DNA from the mitochondria is being used forensically,because each cell has many mitochondria: the mtDNA is a relatively smallmolecule, and some regions of the molecule are very polymorphic WhenDNA sequences are compared, the scientist looks for identities or differences

In the example below, the sequence in (a) is identical to that in (c), but not

by chance (see Figure 11.16)

Variability and Frequency Criteria: The Basis for Frequency Determination in Genetics and Forensic Serology

Human identification depends on two things: (1) characteristics which varyamong individuals, and (2) knowledge of character percentages or frequen-cies Unless the degree of character variation is known, the characteristic is

useless for purposes of identification Consider hair color: describing a person

as having blond hair is very useful identification information in China; it isless useful in Sweden The reason is that a small percentage of Chinese areblond; thus, this information allows us to eliminate a much larger percentage

of Chinese, the nonblonds, than in the Swedish population The process,however, is dependent on knowing that (1) hair color varies among individ-uals, and (2) the percentage of people having different hair colors Thepercentage of each type may or may not differ across different populationssuch as Swedes or Chinese These two qualities — variability and knowledge

of frequencies or percentages — are basic to all identification systems

Figure 11.15 The first PCR-based forensic test was DQalpha: In (A), specimen DNA is placed in a tube along with the DNA replication enzyme (TAQ), buffer, primers, and dNTPs (the nucleotides which will be used to copy the DNA) In stage 2, the tube is placed in a thermal cycler which melts the double helix and then lowers the temperature to permit replication; then, the cycle is repeated Test strips are shown in stage 3; the single-stranded DNA for most of the alleles has been fixed to a nylon membrane which is probed with the amplified specimen DNA This amplified DNA will bind to its complement, which is visualized (B) by coupling biotin to the amplified DNA This biotin is detected by strepta- vidin coupled to a peroxidase which converts a colorless substrate to a dye which is then read as a colored dot Genotypic determination for DQalpha is done by reading sets of colored dots (Figure courtesy of FBI Laboratory.)

A

©1997 CRC Press LLC

The Bertillon System of Identification Provides an Analogy

to DNA Fingerprinting, Although Its Use Was Discredited Many Decades Ago

An identification system using measurements of a person’s head size, rightear size, left foot size, color of the iris of the left eye, and hair color, amongother characteristics, was introduced to forensics by the French anthropolo-gist Alphonse Bertillon in the late 1800s.32 All of the characteristics he used

Figure 11.15 (continued)

Table 11.7 Frequencies of the Various Types of Amplitype PM in Three Populations

Locus Allele Black Caucasian

Hispani c LDLR A 0.25 0.43 0.48

B 0.75 0.57 0.52 GYPA A 0.55 0.48 0.61

B 0.45 0.52 0.39 HBGG A 0.42 0.53 0.39

B 0.26 0.45 0.56

C 0.32 0.02 0.05 D7S8 A 0.66 0.58 0.66

differed among people, and frequencies of the characters were estimated asdata was collected The characters of head height and diameter were com-bined with hair color to produce a composite The forensic utility of thiscomposite description depended, however, on knowledge of character fre-quencies, either the individual character frequencies or the composite fre-quencies Of course, any system claiming individual identification based on

an empirical database would have to measure everyone For example, headheight and diameter measurements of 24 × 22 cm are seen 3% of the time

in a database; 35% of the people in this same database have brown hair Ifthese two characteristics are independent, we may estimate the head size/haircolor composite frequency by multiplying the individual frequencies: 3% (24

× 22 cm head) × 35% (brown hair) = 1.05% If we are correct in assumingcharacter independence, this method of multiplication of frequencies of com-binations of characters will quickly and inexpensively produce identifications

of individuals By adding characteristics, an increasingly rare description isproduced This composite description has both qualities required for iden-tification: variability and knowledge of character frequency

Table 11.8 The HLA DQalpha Genotype Frequencies in Three Populations

DQa, Genotype Caucasian (N = 737) a Black (N = 589) Hispanic (Composite)

a Number of people typed.

b Frequency of DQa types.

Source: Data courtesy of FBI Laboratory.

©1997 CRC Press LLC

Our knowledge of character frequency, however, is predicated on theassumption that these characteristics are really independent In other words,our composite descriptions of uniqueness are based on multiplication ofcharacter frequencies such as head size and hair color; however, multiplica-tion may not be appropriate To answer this question, we could construct adatabase of all composite characters In this database, all entries will be 1divided by the number of individuals studied, or 1/N This empirical method

is fine and may be used to support the contention that the Bertillon fication system, for example, provides individual identification or uniquedescriptions Once two individuals with the same measurements are found,the whole system must be expanded to include new characteristics or aban-doned because the assumption of character independence is only empiricallyjustifiable It is not based on any underlying principles This importantpoint will be clearer when we look at characteristics for which the inher-itance patterns or the genetics are known The Bertillon identificationsystem was abandoned in the 1930s, after two individuals were found tohave the same measurements

identi-Figure 11.16 Surnames may be analogous with DNA sequences: Picking one sequence

as a base sequence, we may describe the changes in the DNA sequence Suppose that we select SMYTHE as the base sequence It is related to SMYTH by deletion of the terminal

E It is related to SMITH by deletion of the E and substitution of I for Y All of the names may be related in this fashion, but the changes required to make these relationships will depend on our selection of a base name or sequence (Figure courtesy of FBI Laboratory.)

The Fingerprint System of Identification: the Most Powerful Identifier of an Individual Since the Beginning of the Twentieth Century

Fortunately another identification system, fingerprinting, had been duced to the U.S in 1902 by the New York Civil Service Commission as ameans of individual identification The Federal Bureau of Investigationadopted the procedure and by 1933 had an operating latent fingerprintsection In fingerprinting, the variables are classified as arch, loop, and whorl;each of these types exists in a portion of the population (65% loop, 30%whorl, and 5% arch) As in physical description or identification, we have asystem based on variable characters of known frequency The basic principles

intro-of identification are the same Knowing that a fingerprint is a loop or a whorldoes not allow us to match another print on file It does allow us unequivocalelimination of all arch prints from further consideration Adding featuressuch as ridge characteristics permits further exclusion For example, if theloop or whorl print also has various ridge characteristics such as dots, bifur-cations, and ending ridges, we may eliminate all prints which do not sharethese features in the same area or position from further consideration Theconsideration of multiple characteristics of a fingerprint is the crux of theidentification potential of this system More and more prints are excludeduntil a unique fingerprint pattern is described In other words, a fingerprintidentification like a physical appearance identification, is a composite wherepositive identification depends on elimination of more and more classes foreach characteristic in the description In the Bertillon system, we first elim-inated all head height and diameter measurements which were not 24 × 22

cm — some 97% of all individuals were excluded Only 3% were included.Fingerprint identification depends on similar exclusion and inclusionlogic; if a print is an arch, then all loop and whorl prints (those which arenot arch) are excluded Both physical description and fingerprinting are based

on character variability and frequency information Fingerprinting has theadvantage that prints may be left at a crime scene, but it shares the disad-vantage that explicit frequency calculations are based on the assumption ofcharacter independence which is, of course, statistically tested in forensic labs

Blood Grouping Offers the Distinct Advantage

of Being Firmly Based in Human Genetics

While blood group and serum protein characteristics share the variabilityand frequency knowledge requirements seen in the use of physical descriptionand fingerprinting, they offer the distinct advantage that the assumption of

©1997 CRC Press LLC

independence is based in the science of genetics Patterns of inheritance forphysical characteristics such as skull dimensions and for fingerprints arecomplicated and incompletely understood On the other hand, the genetics

of ABO, MN, Rh, PGM, and other traits used in serological identificationare simple and well understood This important difference between geneti-cally defined polymorphisms (multiple forms of a character) and those poly-morphism which are not well known genetically is important in light of oursecond criteria for polymorphism — knowledge of frequencies of the variousforms of each character We may employ basic principles of the science ofgenetics to estimate frequencies

Class vs Individual Characteristics: The Cornerstones

of Identification in Forensics that Brings the Value

of Serological Evidence into Perspective when Understood by Laymen

There are three categories of evidence submitted to a crime laboratory: class,individual, and an intermediate category where class evidence approachesindividual evidence Class evidence can be categorized to a specific group orcategory Its rarity or uniqueness is derived from the rarity of the group itself

An example of this is a blood stain categorized as belonging to a member ofthe human race or a higher primate species Most forensic tests cannotdifferentiate between higher primates such as chimpanzee and gorilla andhumans Since there are approximately 6 billion or so higher primates, thisstain is in the class of higher primates, admittedly a large group but onewhich eliminates dogs, cats, mice, fish, etc

Intermediate evidence is class evidence that has characteristics approachingthe uniqueness of individual evidence In terms of serological evidence, theintermediate category encompasses a broad spectrum of tests and procedures

Table 11.9 Phosphoglucomutase Types (Phenotypes) Observed in Samples from Three Countries

Country PGM 1 PGM 2-1 PGM 2 Ireland 74.51

%

23.62% 1.87

% U.S 56.55

%

37.30% 6.15

% Turkey 45.83

The resultant characteristic of an ABO antigen test is technically a geneticphenotype This characteristic allows the examiner to narrow down the stainclass even further than a general category of human or animal stain Comparethe ABO locus (position) on chromosome 9 and the VNTR (variable number

of tandem repeats) YNH24 probe at the D2S44 locus on chromosome 2 TheABO locus has four classes or phenotypes, A, B, AB, and O and some thatare rare There are four common alleles at this locus: A1, B, A2, and O.3 The

Figure 11.17 Various identification systems rely on character variability and knowledge

of character frequencies: Characters which do not vary are useless for identification For example, in searching for a particular person, the information that this person has a head and a chest is not useful, because all humans have heads and chests These are monomor- phic (one form) characters If, on the other hand, we are told that the person we seek is a male, this is very useful information It allows us to exclude all females from our search, reducing the search effort by approximately 50%, because females constitute approximately 50% of all humans (A) In an external physical characteristics description system such as that devised by the anthropologist Alphonse Bertillon, character forms are specified and their frequencies are combined (multiplied if independent) to produce a composite descrip- tion of known frequency In this example, measurements of head height and diameter are taken separately The frequencies of the combined measure are plotted on X,Y,Z coordi- nates As in the head and chest vs male example, identification is achieved by successive exclusion Individuals with the smallest head height and diameter frequency are plotted

in the lower left closest to the origin Identifying a person as a member of this class does not identify the individual; it does eliminate all individuals in all the other classes (B), (C), and (D) are hair color, fingerprint, and DNA fragment identification systems; they are similarly used for identification They differ in their powers of identification, because the number of character states and associated frequencies differ.

©1997 CRC Press LLC

Figure 11.17 (continued)

most an examiner can hope for is to find the rarest type, AB, in this system.This type coincides with roughly 4% of the Caucasian population of the U.S.The VNTR D2S44 genetic locus has over 400 phenotypic classes, the mostcommon present in approximately 15% of some populations; other pheno-types are considerably less frequent Using this D2S44 locus significantlyincreases the informative value of the stain The odds of finding a personwith the same phenotype can range from 1 in 6 to 1 in several thousand.Genes like these clearly fill the requirements of an efficient system for iden-tification They are highly polymorphic and the frequencies are rather low.Individual evidence is at the opposite pole from class evidence Individualevidence itself is in such a rare class, or its individual characteristics are souncommon, as to make it unique A good example of this is a fracture pattern

of a broken piece of glass Every piece of glass that is broken produces aunique and individual fracture pattern If you were to break a glass an infinitenumber of times, a particular fracture pattern would never be repro-duced.5,7,11 Another example would be the arrangement of the nucleotides orbases in a person’s DNA molecules Because of the laws of genetics, yourDNA will never be duplicated again, except in an identical twin Using testsfor different loci such as 4 VNTR loci and 8 to 12 STR loci, the odds offinding the same type in a random population can be as little as 1 in severalhundred million or even billions or trillions

Elimination or Inclusion Is the End of the Journey

in the Forensic Scientists’ Quest for Information

Often the question asked is “How can I eliminate or include person A as thesource of items of crime scene evidence?” This is the issue in most casesinvolving suspects being held or charged with a crime The most importantfactor then becomes selection of the appropriate genetic marker system,which should be chosen to include or exclude the victim or defendant.Forensic serologists have a growing arsenal of genetic markers ranging fromABO antigen typing to DNA/RFLP analysis at their disposal Some geneticmarkers are relatively weak discriminators, whereas others provide extremelyhigh discriminatory power and can approach individualization

A serologist uses a simple statistical test to measure the value of geneticmarkers for individualization This is the power of discrimination (PD) test.11

Two premises must be established The frequency distribution of the types

in the system must be known from population surveys; it is hoped that thedistribution is known in many different populations Next, the genetic

©1997 CRC Press LLC