District Court for the Western District of New York rules on whether evidence obtained using the STRmix method should be excluded from trial (United States v. Pettway)

The Policy

What it does

Holds that evidence obtained using the STRmix method, a probabilistic genotyping software used to match DNA, should not be excluded and that a pretrial hearing is not warranted.


United States v. Pettway is an application of the “Daubert” standard. This standard, deriving from Daubert v. Merrel Dow Parms, Inc., states that a judge must make “a preliminary assessment of whether the reasoning or methodology underlying the expert testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue.”

The Court, after reviewing both sides’ claims, states that the STRmix method of analyzing DNA evidence passes the Daubert standard; i.e., STRmix is a scientifically valid technique for consideration in court. Thus, the evidence should not be precluded, and there is no reason to hold a pretrial hearing on the matter.

While the government notes that this jurisdiction and others have already allowed STRmix evidence, this decision demonstrates the willingness of federal courts to allow this method in analyzing DNA samples in criminal cases.


The Science

Science Synopsis

According to the case, a process known as PCR/STR (polymerase chain reaction / short tandem repeat) was used to analyze the DNA. STRs are DNA regions that have short, repeating units that are usually two to six base pairs (i.e., units) in length. They have become popular for use in forensic science because of their ability to be easily amplified (i.e., reproduced multiple times). Amplification takes place via PCR, a tool used in molecular biology to generate many copies of a particular section of DNA by using only a small amount of the initial DNA.

An additional factor making STRs suitable for DNA analysis is the fact that STR regions vary considerably among individuals, meaning STRs can be used to identify specific persons. Moreover, PCR can even amplify the STRs of a degraded DNA sample, a common finding when looking for DNA evidence, further justifying its use for forensic science.

As the case notes, the PCR/STR process itself is used world-wide and is not under dispute. However, once the DNA from an unknown source has been amplified, it still must be compared to DNA samples from the suspect to see if they “match.” In this case, the relatively new STRmix software was used, a type of probabilistic genotyping software. This class of software uses statistical modeling to “unscramble” crime scene DNA—which often contains the DNA of multiple individuals, or which might contain trace amounts of any person’s DNA—so that unique DNA signatures can be created. Each signature corresponds to one individual. The software then compares these unscrambled DNA signatures to the DNA samples of the suspects, looking to match the DNA patterns between signature and sample. The software presents a likelihood ratio that any identified signature corresponds to a suspect’s DNA sample.

The Debate