Opinion
September 11, 1995
Edward Zizmor for defendant.
Robert T. Johnson, District Attorney of Bronx County (Barbara Chavez of counsel), for plaintiff.
The defendant stands indicted for the crime of murder in the second degree and other crimes. His fingerprint was lifted from a jewelry box in the blood-drenched bedroom where the deceased was found stabbed to death. Upon the defendant's arrest he was wearing boots which appeared to be stained with blood. The boots were seized and DNA testing was performed upon each of the stains and compared with the DNA profile obtained from the postmortem blood of the deceased. The DNA profile found in the bloodstains on each of the boots appeared to match the DNA profile of the deceased and to exclude the DNA profile of the defendant.
For a detailed discussion of the science of DNA forensic evidence, see People v Castro ( 144 Misc.2d 956); People v Wesley ( 83 N.Y.2d 417 [1994]); Springfield v State ( 860 P.2d 435 [Wyo 1993]); and State v Cauthron ( 120 Wn.2d 879, 846 P.2d 502 [1993]).
The defendant conceded that the DNA RFLP testing conducted by the FBI was performed in conformity with accepted scientific methods. The defendant alleges that there was a substantial error committed in the method by which a match was declared and moves for a pretrial hearing to determine limited specific issues with respect to the DNA testing. He claims that extra bands existing in two of the chromosomes tested indicate that there is no match with the deceased's DNA profile. The defendant further alleges that the DNA profile on these two chromosomes, rather than illustrating a match, demonstrates an exclusion of the deceased as the contributor of the bloodstains. Finally, the defendant alleges that it was error to include in the computations the statistical probabilities arising from the chromosomes which excluded the deceased as the source of the bloodstains on the boots. Even if an issue of fact exists as to the alleged exclusion, the defendant urges that only the ceiling principle probabilities be admitted before the jury.
Where the defendant's motion clearly sets forth limited and detailed issues, which are supported by documentary scientific evidence, indicating that questions exist as to the reliability of the tests performed, a pretrial hearing should be conducted to permit the court to resolve the issues presented. This court notes that in People v Wesley ( 83 N.Y.2d 417, supra) the Court of Appeals offered various opinions on this issue. The plurality opinion approved of a hearing being held during trial but prior to the evidence being offered. The concurring opinion opined that the hearing should take place prior to trial. This court deems it crucial for the parties to know, prior to jury selection, if major DNA evidence allegedly linking the defendant to the crime will be admitted at trial. It is extremely important for the parties to have this critical evidentiary question resolved prior to the commencement of the voir dire to aid them in conducting an intelligent jury selection. Secondly, the DNA evidence will undoubtedly be an important issue to be discussed in the prosecution's opening statement and, if desired, by the defense. To leave this issue to a hearing in the midst of trial can result in a substantial risk of great prejudice to one of the parties. For the foregoing reasons, the defendant's application for a limited pretrial hearing to determine the limited issues advanced is granted. It is further noted that the People do not object to conducting a hearing at this stage of the proceedings. (See, People v Ladson, 166 Misc.2d 631 [Andrias, J.].)
This area takes on special significance in view of the massive publicity that has recently been generated by the DNA issue in a highly publicized California murder case.
The hearing consumed only an afternoon session and generated a transcript of approximately 140 pages. (Compare, People v Castro, 144 Misc.2d 956, supra, which lasted about 12 weeks and generated a transcript of about 5,000 pages.)
At the hearing, the People called one witness, Dr. Keith Howland, a Supervisory Special Agent with the FBI. Dr. Howland has been assigned to the DNA Analysis Unit of the FBI laboratory in Washington, D.C. since June 1988. He has interpreted in excess of 7,000 DNA tests. He was subjected to and passed all proficiency tests conducted since his employment with the FBI. He has been declared an expert in forensic DNA analysis approximately 49 times in various State, Federal and international courts, including courts in New York City and State. Dr. Howland was declared to be an expert in DNA forensic analysis without objection. The court finds Dr. Howland's testimony to be trustworthy, consistent and to have the force and flavor of credibility. The defense was permitted, upon their application, to have Dr. Lawrence Koblinsky, a forensic scientist and professor at John Jay College of Criminal Justice, to sit at the defense table to assist in the defense. However, the defense did not present any witnesses.
FINDINGS OF FACT
On March 7, 1994, Dr. Howland conducted DNA RFLP analysis on the bloodstains removed from the right and left boots seized from the defendant as well as the postmortem blood of the deceased. Four genetic loci were examined, D2S44, D17S79, D5S110 and D10S28, and the results were reported. On September 8, 1994, two more genetic loci became available to the FBI, D1S7 and D4S139. DNA RFLP testing was conducted a second time on the same evidence using these recently added loci. Both sides concede that in performing the testing the FBI protocol was followed. It was further established that the FBI protocol conforms to the TWGDAM guidelines. As a result of all the DNA testing, 13 autorads were produced: D2S44 — 2 autorads; D17S79 — 3 autorads; D1S7 — 2 autorads; D4S139 — 2 autorads; D10S28 — 2 autorads; D5S110 — 2 autorads.
The Working Group for DNA Analysis and Methodology (TWGDAM), an association of prominent scientists, has established guidelines for the performance of DNA technology.
After visually examining all the autorads, Dr. Howland offered his opinion that, within a reasonable degree of scientific certainty, the deceased was the source of the blood on each of the defendant's boots and the defendant was excluded as the source of the blood. Dr. Howland also noted that computer sizing measurements were conducted on four of the six chromosome sites: D2S44, D17S79, D5S110 and D10S28. The measurements revealed the following data: (numbers in parenthesis represent the band)
LEFT RIGHT LOCUS 2.5% DECEASED BOOT BOOT DEFENDANT
D2S44 (1) 26.4 1057 1053 1058 780 D2S44 (2) 78.5 3140 3119 3110 1780
D5S110 (1) 77.8 3112 3114 3113 4119 D5S110 (2) 90.9 3639 3652 3650 1152 D5S110 (3) — — — 1083
D10S28 (1) 65.9 2637 2638 2626 2109 D10S28 (2) 72.2 2889 2892 2889 1897
D17S79 (1) 32.6 1305 1292 1297 1493 D17S79 (2) — — — 1265
These measurements confirmed the visual match and led Dr. Howland to conclude that the deceased was probably the source of the blood on the boots and the defendant was excluded as a source of these bloodstains. All the computer-generated sizing measurements fell well within the plus-or-minus 2.5% matching window approved by the FBI.
The issue, as noted by the defendant in his moving papers, became immediately apparent when sections of two of the six chromosomes that were tested revealed extra bands. Dr. Howland testified that this phenomenon, known as "partial digestion", was clearly noted on various autorads in that it caused the production of longer extra bands on a specific site (locus) on the second chromosome (D2S44) and on a specific locus on the seventeenth chromosomes (D17S79). Dr. Howland explained that these extra bands are occasionally observed and have been the subject of scientific scrutiny. These extra bands fall within a designated area and have a known particular average length, with varying longer lengths of base pairs. The extra bands are readily observable to the reader of the autorad. HAE 111 is designed to cut the DNA whenever the sequence GGCC appears in the human DNA. Partial digestion occurs when the restriction enzyme does not cut the DNA at every locus where the GGCC appears on the genome. The enzyme may add base pairs before or after the primary site. The uncontradicted evidence demonstrates that this phenomenon is not an uncommon event. Several scientific principles account for this banding pattern. First, HAE 111 may miss a location if, when the DNA is purified, residual proteins cling to the extracted DNA and masks the recognition site. Second, forensic DNA is subject to degradation which occurs naturally as soon as blood leaves the human body. This may result in the loss of a base pair at the recognition site, causing the restriction enzyme to miss the site. Third, the DNA may have been the subject of a natural mutation at the recognition site, again causing the restriction enzyme to miss the appointed site. Each of these phenomena can cause HAE 111 to miss its recognition site and result in "partial digestion", to wit: extra bands appearing in a DNA profile from a single person. Nonetheless, partial digestion has been studied by scientists. These DNA tests resulted in consistent expected banding patterns of certain average longer lengths of base pairs. There are certain characteristics which are common to all partial digestion bands. Their identification is based on these characteristics and is easily observable on the autorad. On the second chromosome, and in the case of locus D2S44, partial digestion results in a fragment that is, on average, 550 base pairs too long on one side of the GGCC recognition site, and a fragment that is, on average, 1730 base pairs too long on the other side of the recognition site. On the seventeenth chromosome, and in the case of the locus D17S79, partial digestion results in fragments that are, on average, 100 base pairs too long on one end and, on average, 200 base pairs too long on the other end. While this phenomenon occurs most often at the D2S44 and D17S79, it has been observed when sites involved with the fifth (D5S110) and tenth (D10S28) chromosomes are probed. In addition, the smaller fragments generated on sites at the first (D1S7) chromosome are sometimes subject to partial digestion. Although the fourth (D4S139) chromosome is also used in forensic DNA testing, this phenomenon has not been observed at that locus. Partial Digestion: D2S44
The following listing is a guide to interpreting partial digestion in the six probes which are most often used in forensic DNA (D2S44, D17S79, D1S7, D4S139, D10S128, D5S110). The numbers are approximate and represent base pairs that may flank the HAE 111 site on either side of the tandem repeat base pairs and are larger than the primary bands by the approximate number of base pairs (bps) indicated. The third largest number, if present, arises only when there is extensive partial digestion and results when the restriction enzyme misses at both flanking HAE 111 sites. Partial digestion may add base pairs either before the first primary site (1st site) or after the primary site (2nd site) or on both sides of the primary site.
GENETIC FLANKED FLANKED FLANKED LOCI SITES 1ST SITE 2ND SITE BOTH
D2S44 550 bps 1730 bps 2280 bps D17S79 100 bps 200 bps — D1S7 200 bps 400 bps 600 bps D4S139 None noted None noted None noted D10S128 260 bps 1330 bps 1590 bps D5S110 160 bps 320 bps 480 bps
Dr. Howland's inspection of the autorads for D2S44 for both the deceased's blood and the questioned stains on the boots indicated two larger bands on the deceased's DNA profile that were the result of partial digestion. In contrast, there was no partial digestion at that locus for the defendant's blood. The explanation for this phenomenon was that the deceased's blood sample was obtained post-mortem and the bloodstains on the boots had degraded after the blood was spilled from the body of the deceased.
For an excellent discussion of degradation and its affect upon postmortem blood, see People v Ladson (supra).
Most importantly, the partial digestion produced bands where the scientist would expect to find them. At the locus D2S44, one band was 550 base pairs longer than anticipated at one end of the sample and 1730 base pairs longer than anticipated at the other end. FBI protocol permits computer sizing when partial digestion is present in the analytical gel. The sizing measurements showed a band for the deceased at 3140 and a partial at 3595 (+455 base pairs or 95 less that the average size for a partial digestion) and a partial band at 4845 (+1705 bands or 25 less that the average size for a partial). From the sample obtained from the left boot, Dr. Howland found a band at 3110 and a partial band at 4799 (+1680 base pairs or 50 less than the average). On the right boot, he found a band at 3110 and a partial at 4844 (+1734 base pairs or 4 more than average). Visual examination also revealed that the partial bands found in the deceased's blood sample and the boot samples comigrated, an expected result given the matches at the other loci. Dr. Howland noted that a mixture of body fluids could not have produced this type of phenomenon.
The FBI protocol does require a redigestion when the yield gel shows partial digestion. In this case the yield gel did not show partial digestion. This phenomenon became evident only in the analytical gel.
If the FBI assigned its 2.5% window to the partial at 4845 base pairs, that would produce 121.1 base pairs. If that were applied to the partials at 4799 base pairs and 4844 base pairs, these bands would fall within the FBI matching window.
Partial Digestion: D17S79
Partial digestion was also noted at the D17S79 locus. As expected, the partial bands produced band segments that were 100 and 200 base pairs longer than they should have been. The measurement showed a band at 1305 base pairs and a partial at 1508 (+3 base pairs or 3 more than average). From the sample on the left boot, a band was observed at 1292 base pairs and a partial at 1398 (+6 base pairs or 6 more than average). On the right boot sample, a band was found at 1297 base pairs and a partial at 1402 (+105 base pairs or 5 more than average). For purposes of measurement, Dr. Howland assumed that the deceased had a single band at this locus, and did not size the other bands. No partial digestion appeared at the other loci. Applying the FBI binning method to this data, Dr. Howland concluded that the likelihood that the blood on the boots came from someone other than the deceased is 1 in 6 million for Caucasians, 1 in 11 million for Blacks and 1 in 9 million for Hispanics. Using the ceiling principle recommended by the 1992 NRC report, Dr. Howland computed the likelihood that the blood came from someone other than the deceased to be 1 in 161,000.
The partial bands at 1398 and 1402 fall outside the matching window of 2.5% at the 1508 base pair location. That window produces 37.7 base pairs. However, this was explained as the partial digestion for the deceased's blood fell on the 100KB side of the recognition site and the partial on the boot blood fell on the 200KB side of the recognition site. The blood samples on the boots fall well within the matching window for those bands.
Although it could not be ascertained if the deceased was a homozygote at this locus without family studies, Dr. Howland used her value as 2P rather than P squared for the purposes of the Hardy-Weinberg formula for probability calculations.
Computerized measurements were not conducted of the genetic profiles for the D1S7 and D4S139 loci because bands appeared above the 10KB mark on the autorads. FBI guidelines, based purely on policy, forbid the sizing of bands above 10KB. Accordingly, Dr. Howland declared these results inconclusive and did not include them in determining the statistical probabilities of the deceased being the source of the blood. However, Dr. Howland did indicate that all the bands on these two tests appeared to match visually. If they did not match visually, he would have declared an exclusion of both the deceased and the defendant as a source of the blood.
CONCLUSIONS OF LAW
DNA evidence is admissible at trial when a proper foundation has been laid for its acceptance. In People v Wesley ( 83 N.Y.2d 417, 429, supra), the Court specifically made a distinction between the Frye issue and the foundation required for the admission of evidence and noted: "The focus moves from the general reliability concerns of Frye to the specific reliability of the procedures followed to generate the evidence proffered and whether they establish a foundation for the reception of the evidence at trial. The trial court determines, as a preliminary matter of law, whether an adequate foundation for the admissibility of this particular evidence has been established."
The defense concedes that the FBI conducted the DNA RFLP tests in accordance with accepted scientific techniques and their protocol. There is thus no question presented concerning the proper foundation for the admission of this evidence at trial as it has been clearly established. (People v Wesley, supra, at 425, 436; People v White, 211 A.D.2d 982; People v Giomundo, 209 A.D.2d 953 [4th Dept 1994].)
The defense argues that the evidence is inadmissible for three basic reasons: (1) one of the partially digested bands on the D17S79 locus falls outside the FBI's matching window; (2) there should have been a second restriction for the analysis of the D2S44 locus because of partial digestion; (3) the FBI is precluded from relying on bands that fall above 10KB at the D1S7 and D4S139 loci to support its conclusion that the deceased's blood appears on the defendant's boots.
In ruling that DNA evidence is admissible, the Court of Appeals in People v Wesley (supra, 83 N.Y.2d, at 436) noted that: "infirmities in collection and analysis of the evidence not affecting its trustworthiness go to weight, to be assessed by the jury." Ancillary issues regarding integrity of the particular forensic sample from which the DNA fingerprint was obtained and whether the laboratory followed the accepted procedures in carrying out the tests on the particular sample at issue speak to the weight the evidence is accorded and thus are not relevant to the initial determination of admissibility.
Under the controlling precedent as outlined in the Wesley decision (supra), the arguments advanced by the defense do not affect the admissibility of the evidence. Rather, these issues go to the weight of the evidence and may be argued by the defense to the jury if they choose to do so at trial.
In People v Moore ( 194 A.D.2d 32 [3d Dept 1993]), a Cellmark generated DNA RFLP test was ruled admissible where the defendant's blood showed a six band pattern which matched the six band pattern of the questioned sample. However, the questioned sample contained a seventh band that was absent in the defendant's profile. Cellmark explained that the seventh band fell outside its window of measurement and was, therefore, not quantified. The Court held that this type of "ancillary" issue goes to the weight of the evidence and affirmed the trial court's decision to admit the DNA evidence.
This court is required to follow the precedent of an Appellate Division ruling where there is an absence of a contrary precedent from the First Department. (Mountain View Coach Lines v Storms, 102 A.D.2d 663 [2d Dept 1984].)
Although the DNA testing issues herein noted have not been analyzed by any court in New York State, the problems arising due to partial digestion have been somewhat discussed by courts in other jurisdictions. In Springfield v State ( 860 P.2d 435 [Wyo 1993]), the court affirmed the introduction of DNA RFLP forensic evidence. It noted that during the restriction enzyme or digestion phase of the RFLP testing: "Degraded DNA samples can also complicate this procedure. If a DNA fragment is contaminated, this may cause the restriction enzymes to cut the DNA in the wrong places, causing fragments to be different lengths than they ordinarily would be." (Supra, at 440, n 6; see also, Fishback v State, 851 P.2d 884, 887, n 6 [Colo 1993].)
In State v Cauthron ( 120 Wn.2d 879, 897-898, 846 P.2d 502, 511-512, supra), the court affirmed the use of DNA evidence and noted: "Partial digestion of the fragments when applying the restriction enzyme, or its converse `star activity', which occurs when the restriction enzyme cuts in too many places, can also create problems with the autorad * * * While these problems are of concern, they do not require excluding the evidence altogether * * * both the proponents and opponents of a particular test should be able to garner the necessary information to present both sides of the issue to the factfinder."
In the instant case, the measurements which have been set forth clearly indicate that the DNA profile of the deceased fits comfortably within the FBI's matching window. The partially digested bands appear to fall within the area where the scientist would expect them to be. The partially digested bands that do not match have been accounted for based on scientific data. Dr. Howland's uncontradicted testimony undercuts the defendant's argument that the untrustworthiness of the DNA testing process precludes the admission of DNA evidence. His testimony does present questions of fact which may be argued to the jury. However, these questions do not preclude the admission of the DNA evidence.
Accordingly, it is the judgment of this court that the DNA RFLP evidence is admissible along with the statistical probabilities generated by both the FBI binning method and the ceiling principle.