Opinion
J-A25005-17 No. 2014 EDA 2016
01-26-2018
COMMONWEALTH OF PENNSYLVANIA v. ERIC MCCLELLAN Appellant
Appeal from the Judgment of Sentence March 1, 2016
In the Court of Common Pleas of Montgomery County
Criminal Division at No(s): CP-46-CR-0004322-2015, CP-46-CR-0008053-2015 BEFORE: OTT, J., STABILE, J., and STEVENS, P.J.E. CONCURRING OPINION BY OTT, J.:
Former Justice specially assigned to the Superior Court.
I concur with the majority in the result of this matter. There is no question that the evidence presented at trial was sufficient to support the convictions. I write separately to express my concern over the use of the DNA comparison method referred to as "likelihood ratio," and trial counsel's agreement to allow such evidence to be presented without apparent foundation.
Cognizant that probation or parole officers still need reasonable suspicion to search an area, I am not convinced that the search of McClellan's basement was necessary or proper. Nonetheless, I believe that consideration is secondary to the issue concerning the likelihood ratio, which ostensibly linked McClellan to the gun, thereby providing evidence to support constructive possession of the weapon. However, this issue cannot be resolved on direct appeal, because all of the evidence regarding the likelihood ratio was admitted without objection by Appellant's counsel. Therefore, any investigation into this issue must take place in the context of a Post Conviction Relief Act petition.
McClellan's parole officer, Agent Scott Dominick testified he was aware of the basement and had been there once, when he first surveyed the house. However, he had never met or spoken with McClellan in the basement. All prior meetings between the two took place either in the living room or in McClellan's room. See N.T. Suppression Hearing/Bench Trial, 11/30/15, at 59.
Testimony at the suppression hearing/bench trial demonstrated the DNA samples were taken from the gun in question from four areas that were most likely to have been touched by any person: the slide, the grip, the trigger, and the magazine. Traditional DNA testing, where a comparison is drawn between the sample and the defendant, could not positively link McClellan to the weapon. The DNA analysis showed McClellan could not be excluded from having touched the gun in those four areas. I note that when a positive link is demonstrated using accepted DNA analysis, the result is often related in terms of odds in the millions or billions to one - e.g. the likelihood that the defendant provided the sample DNA, as opposed to anyone else, is stated as a factor of a billion to one, thereby supporting a reasonable doubt standard. No such number could be generated linking McClellan to the gun. Initial DNA analysis showed three contributors of DNA in each of the four samples. Instead, the Commonwealth presented evidence of something referred to as the "likelihood ratio."
Thomas Walsh, a forensic biologist for NMS Laboratories, testified there were similarities between McClellan's DNA and the DNA obtained from the gun. However, those similarities were insufficient to provide a positive match. See N.T. Suppression Hearing/Bench Trial, 11/20/2015, at 138.
Likelihood ratio appears to provide a statistical analysis of the likelihood that, in this instance, McClellan was among the three people who provided the DNA sample analyzed as opposed to three unrelated persons having touched the weapon. Further statistical analysis was done by an unnamed source to determine the likelihood ratio of McClellan versus a relative of McClellan having touched the weapon. I find it significant that no DNA sample of any of McClellan's relatives who shared his residence was provided for actual analysis. The likelihood ratio is expressed in terms of tens to thousands of times more likely, rather than millions or billions of times more likely.
Initial DNA analysis showed three contributors of DNA in each of the four samples.
There is no clear explanation in the testimony or exhibits from the suppression hearing of how the likelihood ratio is generated from samples that otherwise cannot link a person to that sample. This is more concerning considering that the unnamed entity that conducted the likelihood ratio calculations was unavailable to explain either its methodology or the scientific/statistical principles supporting that ratio.
A Westlaw search of the term "likelihood ratio" found three cases referring to this method of analysis. They are, Commonwealth v. Trieber , 121 A.3d 435 (Pa. 2015); Commonwealth v. Lyons , 79 A.3d 1053 (Pa. 2013), and Commonwealth v. Foley , 38 A.3d 882 (Pa. Super. 2012). In Trieber , the likelihood ratio was used on canine DNA to establish what breed of dog supplied the DNA sample. Our Supreme Court did not rule upon the methodology because there was otherwise overwhelming circumstantial evidence supporting the defendant's conviction. In Lyons , the method was only mentioned, indicating there needed to be certain calculations done to account for the possibility of co-ancestry in the DNA sample. Finally, in Foley , a panel of our Court stated that if likelihood ratio had been challenged as novel science and subjected to a Frye test, it would likely have passed. The calculations in Lyons were made by a Dr. Perlin, who had developed proprietary software that had been in use in Europe and had helped Britain develop its DNA base. I note that the proprietary nature of the software made it impossible for the specific methodology to be examined. While it is impressive that Dr. Perlin's propriety software has been in use in Great Britain, I would be hesitant to rely upon a secret methodology for conviction. In any event, there is no indication that Dr. Perlin or his proprietary software was used instantly.
I believe this type of calculation was performed by the unnamed entity in the instant matter. --------
In summary, I have concerns regarding the use of likelihood ratio evidence. In this matter, the entity that made the calculations was not identified and there was no testimony as to how the relevant statistical analysis was achieved. I believe prudence dictates that if such evidence is to be used in the future, there must be a record developed demonstrating methodology and the techniques used to arrive at the conclusion.