MATERIALS USED (cells with nuclei)
Blood cells (lymphocytes
Semen, tissue, urine
Buccal swabs, saliva
Pre and postnatal: CVS and amnio, cord blood
Hair follicle cells

TECHNIQUES USED
Southern Blots
PCR

The ABO blood type locus was used in identity testing early on. However, since there are only three alleles (A,B, and O), it is very limited in the ability to identify or exclude persons. Exclusion of relatedness is possible in some cases, for example when the individual in question is type AB and the child or forensic sample is O. Also before DNA testing was possible, cell surface proteins (antigens) were used. These are coded for by the highly polymorphic loci in the HLA (human leukocyte antigens) on chromosome 6. These loci have 10 - 50 alleles per locus but the cell surface proteins are technically difficult to test for, since they require the use of specific antibodies.

The current techniques employ DNA polymorphisms at first using RFLPs (Restriction Fragment Length Polymorphisms), then VNTRs (Variable Number of Tandem Repeals) and STRs (Short Tandem Repeats). DNA fingerprinting requires highly polymorphic markers for which most individuals are heterozygous. These markers are inherited in a simple Mendelian, codominant fashion. A polymorphic locus is one in which there exists a significant variety of alleles in the human population at that locus. The polymorphic markers used in DNA fingerprinting must be "unlinked" independently inherited loci. The loci used are in non-coding regions of the DNA and since non-coding DNA is significantly more abundant than coding DNA there are many such loci to use.

Detection of these polymorphic loci requires digestion of the DNA by restriction enzymes followed by electrophoresis and Southern blotting of the DNA onto nitrocellulose paper, denaturing the double stranded DNA to single stranded, "blocking" with control DNA, and finally, the application of a labeled probe (e.g., radioactivity or hemoluminescense) to the blot, incubation and washing. Probes are specific, known DNA sequences. Multilocus probes can be the VNTR (or other repeat) itself, in which case there are many bands which look like a "bar code" in a supermarket. A single locus probe is a DNA sequence known to be nearby the tandem repeat and unique to that chromosome and region. Southern blots require a rather large quantity of DNA which is in good condition (not contaminated, not digested or degraded, etc.).

PCR has become the method of choice when the proper equipment is available. PCR requires less DNA but theoretically a single cell worth of DNA can suffice and it can even be partially degraded. Amplification of the desired DNA sequence requires knowledge of the DNA sequences of the regions flanking the repeat region in order to construct the appropriate PCR primers. The Taq DNA polymerase (taken from Thermus aquaticus) can only add dNTPs to the primers. About 25 to 40 cycles provide sufficient material. Amplification can occur only over a few hundred base pairs. Nested PCR is used for degraded samples. Multiplex PCR and differently color tagged primers are used when looking simultaneously at more than one locus. The PCR products are electrophoresis in gels. PCR can also be used on coding regions if the allele sequence differences are known (DQalpha with 6 alleles; ABO, HLA). In these, the exon of interest is amplified.

The theory of identification is simple and straightforward: a sample from an individual is compared to close biological relatives or a crime scene sample. In the absence of mutation, the closest relationships are parent-child followed by full sibs who should, on average, share half of all alleles. However they can be expected to lack shared alleles in up to 25% of loci (i.e., 1/2 x 1/2 = 1/4). The Likelihood of relationship or Likelihood Ratio (LR) is calculated to estimate the significance of the biological relationship. LR = A. probability of the DNA typing results if the individual is the person, parent, child or sib of the reference DNA sample divided by B. the probability the putative relationship is not genuine and only due to chance match between two unrelated individuals. B requires the knowledge of allele frequencies in the relevant population. The higher the degree of polymorphism at each locus tested (i.e., the more alleles that exist and the lower the frequency of each allele in the population), the less the likelihood of a "chance" match between unrelated individuals.

Calculation of the Paternity Index. Is the man excluded? Yes, if the bands do not match; No, if they do. There are only two reasons why the DNA profiles would match. Either the person is the person we suspect or there are two different and unrelated individuals who share this genetic pattern. In the latter case we need to determine how often that DNA profile might occur in the general population. Population data bases are converted to tables indicating what percent of the population has a particular band. After the individual frequencies for each gene pair is calculated, they are multiplied together since they are independent events (unlinked loci). Different loci are examined until the person is excluded or until you get to the probability you have determined you need to be certain. Courts use 99% but you can use 99.9 or 99.8% as you decide. 99.9% means 999/1000 or 1000 X more likely to be the person than not. 99% means 100X more like to be the person than not.

References:
DNA Analysis: Medical Genetics by Conner and Ferguson-Smith, 5th edition, Chapter 3
HLA: Color Atlas of Genetics by Passarge, pp.252 - 3.

Back to home page