|Return to the USDOJ/OIG Home Page
Return to the Table of Contents
The FBI DNA Laboratory: A Review of Protocol and Practice Vulnerabilities
Office of the Inspector General
In order to understand the nature of Blake's misconduct and the deficiencies this review identified in the FBI Laboratory's DNA protocols and practices, we first describe in this Chapter the basic characteristics of DNA and the work of forensic DNA scientists. We describe below the physical structure of DNA, testing methods, and the standards that govern DNA analysis.
I. GENERAL PRINCIPLES OF DNA ANALYSIS
All living things are composed of cells, which typically have a nucleus that regulates metabolism, growth and/or reproduction. In human beings, the nucleus contains chromosomes composed of DNA that encode all of the information necessary to produce a complete human body. Chromosomes store information in the chemical structure of DNA much like a book or a compact disk. The nucleus contains 46 chromosomes, two copies of each of the 23 different human chromosomes. One copy of each chromosome is inherited from an individual's mother and one copy is inherited from an individual's father, giving a child DNA characteristics of both its mother and father.
Source: National Human Genome Research Institute, by artist Darryl Leja at
Approximately 99.9 percent of human DNA is the same. Forensic DNA scientists are only interested in the 0.1 percent of the DNA that varies among people. The human traits that result from the variations in this part of the DNA can be obvious, like different eye color or different blood types, but may also be so subtle that only laboratory testing can detect them.
Each chromosome contains many genes, which are the portions of the chromosome that code for personally identifying characteristics, like hair color or eye color. The characteristics of a specific gene, or of a specific location on a DNA strand, is referred to as an allele. For example, if two people both have blue eyes, then they have the same alleles for their eye-color gene. It has been estimated that only 2 to 3 percent of the information in a chromosome is organized into genes. While the function of the DNA between the genes is unknown, scientists currently believe that it does not code for anything. Since it varies widely among individuals, scientists examine the DNA located between the genes to determine a person's DNA profile. Examining this DNA allows scientists to determine an individual's unique DNA profile (except for identical twins), without that profile revealing personally identifying characteristics or medical conditions.
Even though forensic DNA scientists focus their analyses on specific chromosomal locations that vary widely between individuals, it is not necessary to examine every one of these locations to develop a unique DNA profile for an individual. Rather, scientists need only examine enough locations to virtually eliminate the possibility that two unrelated people have the same DNA profile purely by chance. Under current DNA standards applicable in the United States, an individual's DNA profile consists of the alleles present at 13 specified chromosomal locations. Scientists have determined that, in general, when DNA profiles consist of the alleles present at these locations the probability that two unrelated individuals will have the same DNA profile purely by chance is less than 1 in 200 billion. As a result, except for identical twins, examining the 13 locations produces a DNA profile that is essentially unique to an individual. See Appendix 1 (which contains an example of a complete DNA profile).
Law enforcement personnel who submit crime scene evidence for DNA analysis must package and seal the evidence and then arrange for its secure delivery to a DNA laboratory. Upon receipt of the evidence, forensic scientists first determine if the evidence might provide DNA by visually examining it for indications of body fluid stains, and then performing testing to determine whether specific body fluids that might contain DNA are present.
When possible, forensic scientists analyze only a portion of the stains on the evidence and save the remainder in case future testing is necessary. Generally, stains on fabric are cut out of the item and the DNA is extracted from the cuttings. If the stains are on a hard object, such as a knife, some of the dried body fluid is removed from the object with a cotton swab (known as swabbing an item) and the DNA is extracted from the cotton swab. The process used to extract the DNA varies depending on the organic source of the stain and the material containing the stain.
Once the DNA is extracted from the evidence, it undergoes a process known as polymerase chain reaction (PCR), which is also referred to as amplification. This process, often analogized as biological photocopying, allows scientists to make copies of specific chromosomal segments. The amplification process gives forensic scientists the ability to analyze minute DNA samples, and has allowed DNA analysis to become a much more useful tool for forensic scientists. The diagram below illustrates the PCR process:
|Polymerase Chain Reaction (PCR)
[Image not available electronically]
After amplification is complete, the DNA is analyzed using a machine that separates the DNA fragments present in the sample. This process is known as electrophoresis. Special software then measures the length of the DNA fragments, determines the alleles that correspond to the fragments, and compiles a DNA profile for the sample. The DNA testing process is summarized in the diagram on the following page.9
Steps in the Analysis of DNA
As the name implies, short tandem repeat (STR) analysis is a method of determining an individual's DNA profile by counting the number of times a small DNA sequence (short tandem repeat unit) is repeated at a specific chromosomal location. STR analysis consists of three processes: amplification, electrophoresis, and interpretation.
In amplification, extracted DNA is added to chemical reagents and heated, causing the two strands that compose the DNA molecule (they resemble two sides of a "ladder," as seen in the graphic on page 5) to separate. Each of the two strands then can be used as a template to make (or synthesize) a new double-stranded DNA molecule.
The reagents in which the DNA is heated contain markers that identify the starting and ending points of the DNA fragment that is duplicated. The markers also are called primers because they prime (or stimulate) the synthesis reaction. Primers are short synthetic pieces of DNA designed to match the regions of human DNA which are highly variable. As the DNA and chemicals begin to cool, the primers attach to the single-stranded DNA. The primers contain fluorescent labels so that they may be detected by lasers later in the testing process.
Once the primers have bound to the beginning and end of the segment being copied, individual building blocks of DNA from the reagents fill in the rest of the empty spots on the single-strand. See diagram supra at page 7 describing the PCR process.
The heating and cooling of the DNA is accomplished by a machine called a thermal cycler, in which a tray of capped tubes containing the DNA and chemical reagents are placed. The thermal cycler can be programmed to heat and cool repeatedly for specific amounts of time. At the end of many repetitions, millions of copies of the original DNA section are created.
Any DNA present in a tube when the amplification process begins, whether from evidence or introduced through contamination, will be amplified.10 To ensure that the DNA profile generated from the amplified DNA is representative of the DNA from the evidence sample and not from contamination, and to verify that the testing process is accurate, DNA protocols require forensic DNA scientists to analyze a series of control samples. For each batch of samples processed, at least one positive control, one negative control, and one reagent blank are analyzed along with the DNA samples. The positive control tube contains the reagents necessary for amplification plus DNA from a source for which the DNA profile is known. Since the scientists know the correct test results for the positive control, it allows them to determine the accuracy and performance of the amplification and analysis processes. The negative control tube contains all of the reagents used for amplification. The reagent blank contains all of the reagents used to process an item of evidence from extraction through electrophoresis. DNA from the evidence is not added to these controls, though their contents are amplified. The purpose of the negative control and the reagent blank is to reveal any contamination that is present in the reagents or introduced during the testing process.11
TYPES OF DNA CONTROLS
|Positive Control||Reagent Blank||Negative Control|
|Material Tested||Amplification reagents and known DNA||All reagents||Amplification reagents|
|Reveals||Accuracy and performance of the amplification and analysis processes||Presence of contamination introduced at any point in the analysis process||Presence of contamination introduced during the amplification process|
After the DNA has been amplified, the newly formed DNA fragments are sorted according to length (i.e., number of short tandem repeats) using electrophoresis. In general, electrophoresis is performed by adding DNA to one end of a piece of gelatinous material which contains tiny holes that allows the material to function as a molecular sieve. An electric current is applied across the material, causing the DNA fragments to move. Since it is easier for smaller fragments to move through the material, the smaller fragments move farther than the larger fragments. As a result, at the end of electrophoresis the DNA fragments are sorted by size. The size of the DNA fragments is determined by comparing the distance each fragment moved to the distances moved by the fragments of known size. The results of electrophoresis are illustrated in the following graphic.
|[Image not available electronically]|
The principles described above also apply to capillary electrophoresis, a form of electrophoresis employed by the DNAUI. Its distinguishing characteristic is that the electrophoresis occurs inside a capillary tube (a very thin glass tube, comparable to a human hair) with a sieving material inside, rather than on a piece of gelatinous material. Capillary electrophoresis is an automated process that analyzes many DNA samples and requires minimal involvement by DNA scientists after the initial set-up procedures are completed. These procedures include cleaning and calibrating the electrophoresis machine and preparing the amplified DNA for analysis.
To prepare amplified DNA for capillary electrophoresis, the DNA scientist:
Once the tubes are sealed, the rack is ready to be placed on the capillary electrophoresis machine. A sample list is prepared which identifies the location of each sample on the rack and makes it possible for the machine's computer to locate a specific sample. An injection list is also prepared which tells the computer the order in which the samples are to be analyzed. The capillary electrophoresis machine has a probe that punctures the soft rubber caps on the tubes and withdraws a specific amount of sample. The sample is drawn up into the capillary tube (referred to as injecting the sample) where the electrophoresis is completed.
As mentioned previously, the primers used during amplification contain fluorescent markers that allow the DNA fragments to be detected by lasers. The manufacturer of the capillary electrophoresis machine has developed proprietary software to display the test results and to aid in their interpretation. Using this software, the capillary electrophoresis machine determines the size of the DNA fragments in a sample based on the information detected by the lasers. The machine and the software then represent the lengths of the various fragments as peaks on a graph as illustrated on the following page:
GeneScan® View: raw data for a Positive Control (9947A) prepared according to protocol.
Peaks depicted in red originate from the internal size standard added to each sample.
The proprietary software has two components, GeneScan® and Genotyper®.14 Data viewed in GeneScan®, as appears above, is the raw, unanalyzed, collection data that reflects everything the laser detects, including interference that is common in electrophoresis instruments (Genescan® data). Genotyper® allows forensic scientists to take GeneScan® data and display it in a format that conceals background noise and peripheral information, and to focus their review on the results of the control and evidence samples. An example of a Genotyper® display is presented on the following page:
|Genotyper® View: Profiler Ladder |
with Positive Control Allele call
[Image not available electronically]
Information collected during these analyses is used to assemble the DNA profile. As mentioned previously, two points of reference are used to help the software as it determines the lengths of the DNA fragments detected during electrophoresis: 1) the GeneScan® software uses the internal size standard, which contains DNA fragments of known sizes; and 2) the Genotyper® software uses allelic ladders as a point of comparison for the designation of the number of repeats in the DNA sample at particular chromosomal locations, since the peaks within the allelic ladder correspond to known fragment lengths at those locations. The DNA Examiner then works with the Genotyper® graphs, similar to the one above, looking for any peripheral information that should be considered. Unless contamination is detected or other complications disrupt the testing, the Examiner then documents what the allele values are at each of the chromosomal locations analyzed (usually 13 chromosomal locations are examined), which, once compiled, constitute a DNA profile. See Appendix 1 for an example of a complete DNA profile and the corresponding GeneScan® and Genotyper® graphs.
II. STANDARDS GOVERNING FORENSIC DNA ANALYSIS
The creation of national standards for DNA analysis played a pivotal role in establishing the integrity of the DNA testing process. In addition, by adhering to these standards, DNA laboratories, including the FBI's DNAUI, have been able to attest to the validity and reliability of their DNA testing results.
Forensic DNA laboratories, particularly those participating in the FBI's Combined DNA Index System (CODIS),15 have relied upon three primary sources of operational standards since the first forensic DNA laboratories were established in the late 1980's: 1) the Technical Working Group on DNA Analysis Methods (TWGDAM); 2) the DNA Advisory Board; and 3) the FBI's National DNA Index System (NDIS) program office.
TWGDAM was one of several technical working groups sponsored by the FBI. The goal of the working groups was to improve communication between the various scientific disciplines and to build consensus within the federal, state, and local forensic communities. TWGDAM was established in 1989 with representatives from 12 federal, state, and local laboratories, and focused specifically on the development of forensic DNA methods. Later that same year, TWGDAM developed and published in the Crime Laboratory Digest16 a set of quality guidelines for forensic DNA laboratories.17 TWGDAM expanded these guidelines in 1991 and in 1995.18 In addition, TWGDAM worked with the National Institute of Standards and Technology (NIST) to develop model reference material that laboratories across the country could use to gauge the reliability of their equipment and DNA testing processes. In January 1999, TWGDAM was renamed the Scientific Working Group on DNA Analysis Methods (SWGDAM),19 and in that capacity produced additional guidance for the forensic community, including guidelines for data interpretation, training, quality assurance, and health and safety audits.
While no formal legal authority was granted to TWGDAM and SWGDAM, the guidelines they produced were accepted by the Laboratory Accreditation Board of the American Society of Crime Laboratory Directors as the benchmark for DNA laboratory accreditation. Further, when Congress authorized the creation of CODIS in the DNA Identification Act of 1994,20 it provided that the guidelines issued by TWGDAM would be deemed to be national standards until the FBI issued its own standards pursuant to the Act.
The second source of DNA standards is the FBI DNA Advisory Board (Board). In the DNA Identification Act, Congress required that the FBI establish an advisory board to develop national quality assurance standards governing all CODIS participants.21 As a result, the FBI established the Board, which was formally constituted on March 10, 1995.22 Its members were appointed by the FBI Director based upon nominations from a variety of forensic and science organizations,23 and included forensic scientists from state, local, and private forensic laboratories; molecular and population geneticists; a NIST scientist; a quality control specialist; an ethicist; and a judge. The Board's mission was to develop and revise, as necessary, standards for quality assurance, including proficiency testing standards for laboratories and analysts that examine DNA. The Board members acknowledged that TWGDAM had begun this work and that the Board should build upon it.
The Board fulfilled its mission with the submission to the FBI Director of quality assurance standards for two types of DNA laboratories:
Amendments to these standards must be approved by the FBI Director. Recommendations for changes can be requested through SWGDAM.
The third source of DNA standards is the FBI NDIS program office, currently within the Laboratory Division's CODIS Unit (see the organization chart on page 24 for the placement of the CODIS Unit within the Division). The NDIS office has issued programmatic rules that govern the exchange of information for NDIS participants and has established standards for the submission of DNA data, collectively referred to as NDIS Requirements.
At present, three sets of standards govern the DNA activities of the DNAUI: 1) Quality Assurance Standards; 2) NDIS Requirements; and 3) Accreditation Standards. These standards are interrelated: to comply with the Quality Assurance Standards, a laboratory is supposed to pursue accreditation actively; to become accredited, a laboratory must demonstrate compliance with the Quality Assurance Standards; and to become a participant in NDIS, a laboratory must demonstrate compliance with both the Quality Assurance Standards and the NDIS Requirements. We describe each of the standards below.
Quality Assurance Standards consist of two sets of standards: 1) Forensic Standards that govern the activities of DNA laboratories that analyze crime scene evidence, and 2) Offender Standards that govern the activities of DNA laboratories that analyze samples from convicted offenders. The Forensic Standards contain 155 requirements organized under 15 headings, and the Offender Standards contain 136 requirements also organized under 15 headings.24 For complete versions of the Forensic and Offender Standards, see Appendix 3.
The key categories of requirements addressed in the two sets of Standards, which correspond to section headings in the Standards, are the following:
NDIS Requirements are found in the Memorandum of Understanding (MOU) signed by the FBI and each NDIS participant. The MOU requires that signatories comply with general requirements already established (i.e., federal legislation, the Forensic and Offender Standards) as well as requirements specific to the national index that accompany the MOU in three appendices: NDIS Responsibilities (Appendix A); NDIS Data Acceptance Standards (Appendix B); and the NDIS Procedures Manual (Appendix C).25
The primary accreditation or certification entities for forensic and offender DNA laboratories are the American Society of Crime Laboratory Directors - Laboratory Accreditation Board (ASCLD-LAB) and the National Forensic Science Technology Center (NFSTC). Both groups draw upon the requirements set forth in the Forensic and Offender Standards for their evaluation of a public DNA laboratory's operations.
III. ACCREDITATION AND STANDARDS COMPLIANCE
While TWGDAM/SWGDAM and the Board were pivotal in creating standards for DNA laboratories, they lacked the means to enforce them. To compensate for this shortcoming, the Board adopted an "Accreditation Premise" which set forth the Board's expectation that standards compliance would be assured through the process of accreditation. Accrediting organizations would need to adopt and hold laboratories accountable for compliance with the Board's standards. The Board acknowledged that a weakness in this approach was the lack of any enforceable requirement that laboratories be accredited, even for CODIS participation. In an attempt to address this problem, the Board passed a resolution in February 1999 stating that unaccredited laboratories should seek accreditation "with all deliberate speed." In addition, this language was used in the preface to the Forensic and Offender Standards to emphasize the importance of accreditation.26
Compliance with DNA-related standards is an issue previously examined by the OIG. In 1999, the OIG performed an audit of CODIS to determine the extent of state and local CODIS participation and to verify compliance with the FBI's quality assurance standards and national index requirements.27 In the report summarizing its findings,28 the OIG explained that the FBI's practice at the time of audit fieldwork was to allow CODIS and NDIS participants to self-certify their compliance with the Quality Assurance Standards and with NDIS Requirements. Because the OIG believed this system of self-certification posed a high risk of undetected noncompliance, the OIG undertook compliance testing of various CODIS participants and subsequently identified multiple instances where the participants were not fully complying with national standards. In addition, while the OIG noted that all audited laboratories had complied with the Forensic and Offender Standards' annual audit requirement,29 weaknesses were noted with some of the external audits: 1) audit findings were not binding on the laboratories (they could disregard them if they wanted); 2) although accreditation and certification agencies had the authority to ensure a laboratory took appropriate corrective action, accreditation or certification audits did not typically focus on compliance with the quality assurance standards; and 3) laboratory audits were not always performed consistently. From these observations, the OIG recommended that the FBI develop and implement a process that would ensure that laboratories resolve all deficiencies noted during the external audits.
In response to the OIG's findings and recommendations, the FBI developed a new operational procedure, called National DNA Index System (NDIS) Review of External Audits, which provides for the formation of several NDIS Audit Review Panels. Each panel consists of four qualified or previously qualified DNA examiners or analysts selected from the FBI and state or local laboratories, with the chief of the FBI Laboratory's Quality Assurance and Safety Unit serving as chairperson. All panelists are required to have completed successfully FBI quality assurance audit training. Under the new procedure, NDIS participating laboratories must forward to a review panel, via the custodian of the NDIS database, a copy of their external audit report, their response to the report, and corrective action plans that address the audit report recommendations. The panel reviews the audit report and related documents to determine if all findings and recommendations have been addressed adequately and/or resolved. If the audited laboratory does not respond to clarification requests by the panel, does not resolve an audit recommendation, or is determined to be non-compliant with the quality assurance standards, a corrective action and conflict resolution process can be invoked. A laboratory's failure to resolve a panel's concern can result in the termination of its access to NDIS.
In addition to these compliance procedures, the FBI created a standardized DNA audit guide (Guide) with input from the Board, ASCLD-LAB, and NFSTC to ensure that auditors of local, state, and federal DNA laboratories are thorough and interpret the Quality Assurance Standards consistently. The FBI offers Guide training for auditors, including those representing accrediting and certifying organizations such as ASCLD-LAB and NFSTC. For an audit to fulfill the Quality Assurance Standards' external audit requirement, it must be conducted in accordance with the Guide and by an auditor trained in its use. However, as this report details, even with these precautions, internal control weaknesses are not always uncovered in quality assurance audits. In fact, weaknesses in DNAUI procedures and protocols allowed a technician routinely to disregard required steps in the analysis of DNA, even while the Unit received clean audit reports from both internal and external auditors and while the Unit was accredited by ASCLD-LAB.