DNA Statistical Evidence and the 'Ceiling Principle' Science or Science Fiction?

• Author: Major Douglas A. Dribben
• Pages: 02

In law, the man oflke future is the man ofstalislics. Oliver Wendell Holmes, J?' (1807)

I Introduction

Since 1086, prosecurors and defense attorneys have had a pou erful weapon to aid them m determining the identity of the perpetrator of a cnme.l The forensic use of Deoxpbonucleic Acid (DNA) permits absa-lute exclusmn of a defendant from the group af possible pelpetratars. thus preventing the innocent from conrwtion andpossible Impnsonment. Alternatively, it can protide powerful circumstantial evidence that rhe defendant and the perpetrator are one and the same and help ensure that the guilty are brought to jueuce.

DS.4 e\idence is compnsed of two elements the presence or ab-sence of a"match between the suspect's DSA and the evldentiary sample found at the cnme scene, and the relevance of this match The admission of this endence can take three forms: exclusmn of all the DNA emdence. admission of the issue of a match alone, or admission of both the match and its relevance.

Most state and federal courts hare adrmrted DNA endence in one

form or another. With the demise of the Fwe2 and Fwe-baseds standards of admasibdity (in federal courts and couw-marrial), little or no chai-lenge remains to admtung endence of a match between the evidentiary sample and the defendant's DNA m all federd (mciudmg military) and most state courts.' Ths emdence can, and has, passed scrutiny under the Fedma1 Rules of Evidace (FRE)

However, a controvemy has arisen over the scientific basis used to admit ewdence demonstrating therelmance of arnatch between the DNA of the suspect and the evidentiary sample This eudence usually 1s presented as a statistic-the probability of this match occumng at random from someone other than the perpetratmi This probability usually is eutraardnattiy small, often as law- as one m a million or less. This e m dence is damning in the eyes af the july, and defense attorneys and their expens tly hard to prevent its admissibility.

A nev method of calculating ths statistical evidence was created m response to this controversy. This method, called the "cellmg pnn-cipie," IS unduly cmsewative and aperates to greatly increase the probabihnes calculated by most United States DNA laboratones. Under the guise of science and the cloak of respectabihty promded by 11s sponsor-the Nahonai.4cademy of Sciences-tlus method found its ray into many recent decisions

Ostensibly based on science, this method enters the counroom under the auspices of the rules of ewdence governing admissibility of scientific evidence. Yet the methodlacks ascientific basis and its admis-smn contradicts the pnncipies undedpng the applicable federal rules of emdence The results of chis new method of calculating DSA statistical eiidence may create a reasonable doubt as to the identity of the pelpetratar.6 At the least, they greatly reduce the effectreness of DNA em dence and increase the likelihood of confusing the factfinder.

*Foe

Y United Sfafes, 293 F 1013 (D C Ca 1923)lSer.eg .Peoplev Casfro 646YY S2dQS6(Sup Ct 1989) (addmga1equrement thaffhe laboraton c~mpliaifhpl.operproceduresm conducting DYAL~JIbefore endence

is admissible)

'Daubell v MerreLl Doa Pharmaceuncds, lnc , 113 5 Ct 2786 (1993). Urd on m-mond 43F3d1311191hCir 1991)

. .

This article contends that statmcal endence calculated using the "ceiling principle" (1) IS not based on any scientific theow or body of howledge, (2) grosslyoverstates theprobabilivofarandomDNAmatch. and (3) when introduced into endence alongside or m place af rhe sratistical midence calculated usmg the traditional "product mle,". 1s likely to confuse or mislead the facrfinder, thus creating doubr as ro guilt where doubt otherwise would not eust P;ut I1 of this m~cleprmides a brief aveniew of the process of DNA analysis. Part 111 surveys the history of DNA eiidence in .Amencan courts Part IT addresses the contiovers) surrounding admission of DNA e\idence Part V examines the histoly behind the "ceilmg plinciple" and Its xientitic underpinnings, if ani Part 11 examines the relationship between the "ceiling principle" and the rules of evidence Pan \'I1 conrains the conclusion and recommendations

11. DNA.halys~

Organisms reproduce by transmitting genetic information from gen-eration to generanon \iia the DNA malecule, whch contans genetic codes that determine inherited charactenstics.' In humans. DNA IS conraned m forty-six chromosomes one pmr of sex chromosomes and tnenrytwo pairs of autosomes ' Dunng reproduction, the father's sperm and the mother's mvm each proiide half of an mdwldual's DNA

Geneticists are now able to isolate human genes h h t genes are inrolred in determining rhe structure and function of cells However, some genes hare no apparent function IL These apparently functionless genes exhibit tnde urnations among mdinduals and sere as the basis behind DNA analysis ' I

A 77te Compositmi 0JD.U

DSA is the basic building block of all hwng cells Found pnmanly m che chromosomes wthm the nucleus of all human body cells (except red blood cells).13 the DSA molecule itself 1s composed of two srrands

froin E U ~ ~ P C E

ma? be compared Id at 4

mteItuined in 8 spiral or double-helix formatian (resembling a zipper)."

Each strand contains four different nucleotides, or bases, repeated hundreds of thousands of times. These bases are deaxyadenosme mano-phosphate (A), thgrmdme monophosphate (T), deoxycytAdine monaphosphate (C), and deoxyguanosine monophosphate (G). The bases assoc~-ate mth each other m cemn ways: T on one strand of DKA ulll only band wth A on another strand likenwe, C will only bond with G. However, there are no hmds to association between the bases an the same strand of DNA. Each association between two bases LS hahn as a base pair.'j Consequently, a sequence of DNA molecule may look like.

A T G C C G A T G C A T A G T C A C G T A G C T I I I I I I I I 1 , I I h ~ l

I I I I I I 1 I I

T A C G G C T A C G T A T C A G T G C A T C G A

Because of these associational properties, If the sequence of one strand of DNA is horn, the sequence of the other strand can be deterrmned quite easily.ln

There are over three billion base pars m each strand of human DNA cantamed m each of approxlmately ten tnllion cells m the human body.'> The base pur arrangements wthin the chromosomes form genes Genes help detemune such charactelistics as whether an indi5iduai has blue or green eyes. Alternate forms of genes, such as the "blueeye" and the "green-eye" gene are called alleles. Each human ailele caniains from one to 2000 kilobase pain, or Kb

Most of the DSA in humans is the same from one penan to another. .4n indimduals DNA vaies, however, at approximately three mii-lion sites, or loci.Le These differences-called "polymmorphisms'-accur at discrete loci xithm the genes along the DS.4 strand and exhibit B high degree of venation among individuals Geneucists have discovered that fragments of DNA are repeated many Umes at these sites, mth the "ma-tion occuning in the number of times the sequences are repeatdin The

14SeeznJra Appendu(Fwre L~sadiagrunofrheDNArnolecde)e Because of the large number of base p a s m each allele, DNA sample sizes corn.

F Samuel Baeehtel, A Pnmer m the Methods Csed zn the T w m g oJD.Y.4 15 rnonlyarereferredfain KAohasoa(l(b). or onethouand base pan

CRIME LIBORATORIDIC 3 (1988)l i d lsPeoplei Cmifro 545UI S26965,966(Sup Cr 1989)

"Expertseinmale that at leal me heseperfhousand jmes berueen indrwduali

D K Cooper et al An Estimate oJCnigur DYA Sequence Hri~~o~ygos~f~in the mman

Genome, 68 HLI G~\incs201 ZG6 (1886) 9aForexampl~ inthe sewence

~-C-T-GA.T-G-A.T-GA-T C-6-A-A-T-GA.T.GA.T-T

thepeneiGA.Tiirepeated threetimes alone lo~monandluice at another

vanations in number of the base series repeats are referred to as "van-able number of tandem repeats." or iNTRs

.An indix?duai has at most tho alleles at any one locus-one inherited from the father and one inherited from the motherof these loci hate up to one hundred different allelesphic loci form the basis of DNA identification.

B The ?beow ofD.\A

The DNA within a person's cells 1s )dentical regardless of the type of cell However. no two people ham exactly the same DSA except identical twins ?- These two precepts form the basis of DN.4 analysis. Because of them, DSA from a suspecfs blood may be compared TO a semen sample from the crime scene to determine the identit) of the perpetrator

Cornpalison of DNA samples LS much like cornpansons of a panial fingerpnnt.The humanDNAemuchtoolargetocampareinItsentirety'~ Therefore. only a small portion LS andyed for forensic purposes

If one strand IS knovn the other can be readil) determined due to Its complementary bonding propelties This LS the hem of DNA analys~ The companson is performed by separating the helical molecule into its two component strands and breaking the strands d o m into smaller fragments. Then, a fragment from a strand of the DPA from one source may be comparedtoafragment fromastrandoftheDNAframanothersource If the Dlr.4 LB Identicai, the campiementq fragments will band, If not. no banding will occur. Because the fragrnenrs bond only with their COW- terpm fragments, bonding indicates ihat the two samples themselves match at the points compared.

So matchpra~,desconclusi~-epraofthatthesuspect isnot the cnm-

Bofhpartnfscanparsonthesame genetofheiroiispnng

G ~ E T I C

Srrwri supra note I a1 12

Exceutfors~emicellsandova shlcheachconmn exactlvhallrhe DIAfoundin the olhercel1;the dliferencesbetneen DXhn dlflenngfipesof c& can adsbe defected through speclflc and dewled laboram" testing There minor differences are not derecrable usmethe D \ l a n d i i ~ methods dircusiedmthisarticle Id at42

nal (if they were the same person, the samples should match evewhere, including those portions under examination). A DNA match provides powerful, although not conciusive, evidence that the suspect ( o r b identical tmn, If one exists). prmided the ewdentiruy sample Although the area under exarmnauon matches...