The proliferation of forensic DNA testing and databases by state and local jurisdictions was recently reported by Joseph Goldstein in The New York Times.1 For years, the FBI’s combined indexing system (CODIS) has maintained three national registries of human DNA records. One is a registry of convicted felons, the second is for evidence from crime scenes, and the third is of DNA of unidentified persons.
However, law enforcement is responding to needs at the local level and to incentives to create their DNA databases with much larger scope. The zeal shown by local officials should be balanced with the need to make sure the evidence is analyzed accurately and is available to both the prosecution and defense for critical review. America’s legal system will not survive if it is widely perceived as delivering biased justice.
Local programs differ in scope. Most examine DNA samples from crime scenes. In one case, a doorknob was sampled in a residential burglary, and then DNA samples were requested from the residents of the home. All of the samples were placed in the local registry. There is no way that information can be removed. Worse, it may be shared with other locations without the knowledge or approval of the donors. Simply being listed in a registry may precipitate further questions in a traffic stop, etc.
Denver maintains a local database. The district attorney uses a software program to identify relatives of a person with DNA collected from a crime scene. Potentially, this could be submitted to a grand jury to question family members about the identity and location of other family members, which would probably include the person with the target DNA. Failure to cooperate could potentially lead to imprisonment for failing to cooperate with a grand jury.
In 2013, analysis of DNA is probably the most golden of all the gold standards in forensics analysis. Each of us is different, and the difference is programmed by our unique genetic code. The exception is identical twins and triplets, who share identical genomes.
Still, for identification, many, if not most, think that each of us is uniquely defined by our DNA. DNA may be more definitive for identification than our fingerprints. Several test modes are used to characterize the DNA of a particular individual, including variable number tandem repeats (VNTRs); short tandem repeats (STRs); and Y chromosome STR, which focuses on the Y (male) chromosome in semen.
Even with this, the National Academy of Sciences, in its 2009 report, warned that “errors do occasionally occur.” These are usually traced to differences in interpretation, particularly when the sample is a mixture; failure to process the samples correctly; laboratory contamination; low sample amount; and differences in test protocols.
Plus, DNA testing is not infallible. For example, recall the confusion that resulted when it was reported that chronic fatigue syndrome (CFS) was related to human retrovirus XMRV in prostate tissue. There did not seem to be a biochemical explanation. Ultimately, the correlation was traced to a common primer DNA sequence with 99% homology. Initially this was interpreted as having a common origin or connection. Further work has shown that the conclusion is probably not valid.
Advances in molecular biology will probably uncover other ambiguities that may affect the validity of a particular DNA test. For this reason, the Innocence Project recommends that the DNA evidence be preserved for as long as the person is subject to incarceration, parole, or registration. Further, all records should promptly be destroyed in the event of exoneration or pardon.
Laboratories entering DNA profiles into CODIS must meet accreditation and use validated protocols. Proficiency testing is mandatory. Analysts are required to have at least a bachelor’s degree. The technical leader of the lab must have advanced training.
FBI reporting guidelines require: 1) a description of the evidence, 2) a list of the loci analyzed, 3) a description of the test protocol, 4) results, 5) conclusions, and 6) an “interpretative statement” concerning the inference supported by the analysis.
The report by Goldstein1 is disturbing since there is no assurance that the local databases have been gathered from trustworthy sources. Trustworthy sources include labs that collect and report results according to validated protocols. The staff also needs to be trained and motivated to deliver accurate results in a believable, impartial manner. This is one more example showing the need for a national program as proposed for a Forensic Laboratory Improvement Act.2
Errors in reporting in the forensics lab
The forensics laboratory has a unique requirement to make measurements and report results accurately. The lab can also offer interpretation of the results, but opinions from the technicians need to be recognized as opinions, and these are usually based upon experience. The laboratory may proffer a set of statements that conclude, for example, that the defendant was present at the scene and/or had fired a gun:
- “Gunshot residue was found on the defendant’s gun hand. This is consistent with the defendant having fired a gun recently”
- “The fingerprint of unburned powder on the gun matches the residue on the suspect’s shooting hand”
- “The fingerprint of powder is a rare form of black powder that was last manufactured 27 years ago, and is not in current production and rarely seen today”
- “The gunshot residue indicates that the defendant had fired a similar or the same gun within the last XY hours.”
Analysis of DNA with quantitative PCR
DNA analysis with rtPCR (real-time polymerase chain reaction, or quantitative PCR) is exceptional detection sensitivity—detection of a single molecule is possible. Analysis of specimens obtained from motel rooms or steering wheels of rental cars could wind up in the databases, with the wrong attributes or origin. Indeed, the article in Ref. 1 depicted a half-page picture of a man swabbing the steering wheel of a car to collect a sample for DNA identification. A match to the car may be all a prosecutor needs to elicit a confession from an innocent person who had the misfortune to have used the car days or weeks earlier, such as a parking attendant. The Innocence Project is working on hundreds of cases involving the role of modern DNA analysis in overturning mistaken convictions.
Laboratory contamination is usually among the most often cited sources of error, particularly false positives. A false positive can erroneously place the suspect at the scene of the rape or murder, etc., simply if that suspect’s DNA is found in the assay that purports to have only DNA recovered at the crime scene.
Laboratory contamination is a common problem. A simple Google search on “laboratory contamination” lists many examples describing how contamination can creep in and lead to incorrect results. Some of the most interesting cases deal with contamination of heavy metal samples. For example, mercury can be reduced to Hg0, which makes it soluble in plastic containers. Ionized mercury can be retained on the surfaces of glass by ion exchange.
Several articles discuss contamination of cell cultures with microbes. One talks about infection of inoculated specimens with live polio virus. Many examples read like a murder mystery. Few of the articles are recent, however; laboratory contamination is not a topic with a high impact factor and is hence avoided by the editors of “white” journals.
One reference presents a flow chart for radioactive contamination. It lists pipetting as a likely source of contamination. Glass pipets are difficult to clean. Air-driven pipets can lead to contamination since, during the aspirate phase, the velocity of the load cycle, including repeated cycles of aspirate and dispense, can create aerosols that contaminate the pipet body. This can transfer to subsequent samples. Overly forceful dispensing is also implicated in creating an aerosol that can spread contamination to the surrounding area during dispense. Think this is crazy? Why do we have Biosafety Level 4 (BSL-4) facilities?
The future of forensic DNA analysis
Some will be quick to point out that forensic DNA analysis is based on old technology that looks for variations, often simple repeats, in specific locations. It is not whole genome sequencing (WGS). But, the price of WGS is declining by a factor of 10 every five years, so we can anticipate that soon huge databases of WGS of Americans will be compiled, primarily for precision medicine. Will law enforcement have access to these records? You can bet they will try.
- Goldstein, J. Police Agencies are Assembling Records of DNA. The New York Times, June 13, 2013, p 1A.
- Stevenson, R.L. The Time has Come for a Forensic Laboratory Improvement Act; http://www.americanlaboratory.com/913- Technical-Articles/128817-The-Time-has- Come-for-a-Forensic-Laboratory-Improvement- Act.
Robert L. Stevenson, Ph.D., is a Consultant and Editor of Separation Science for American Laboratory/Labcompare; e-mail: firstname.lastname@example.org.