The Five Steps of Forensic Genetic Genealogy

A hobby genealogist already knows the central truth of family-tree research: one record rarely settles a relationship. A census entry can be wrong. A surname can change. A tree can copy another tree’s mistake. DNA can point toward a family without naming the exact person. Sound conclusions come from convergence — records, relationships, geography, chronology, and genetic evidence fitting together without forcing the answer.

Forensic genetic genealogy, often shortened to FGG or IGG, uses much of that same reasoning. It combines a dense DNA profile with genealogy research to develop an investigative lead. But the setting changes the standard. The unknown person may be an unidentified decedent or the possible source of crime-scene DNA. Privacy interests are broader. Chain of custody matters. Evidence may be limited and irreplaceable. A mistaken branch does not merely produce a wrong ancestor in a private tree; it can direct police attention toward an innocent living person.

That is why the cleanest way to understand forensic genetic genealogy is as a five-step process. Each step narrows uncertainty, but none of the first four proves identity. The genealogy develops a hypothesis. Conventional forensic and investigative work must test it.

Step 1: Build a dense SNP profile from the forensic sample

The process begins with biological material held in a forensic case. In an unidentified-remains case, the sample may come from bone, tooth, or preserved tissue. In a criminal investigation, it may be DNA reasonably believed to have been left by the unknown contributor at the scene. Before genealogy begins, a laboratory must determine whether the sample can support additional testing and whether consuming some of it is justified.

Traditional forensic DNA work usually focuses on short tandem repeats, or STRs. STR profiles are highly useful for direct comparison and for searching law-enforcement systems such as CODIS. Genetic genealogy uses a different kind of profile: hundreds of thousands of single-nucleotide polymorphisms, or SNPs, distributed across the genome. The Department of Justice’s public policy describes forensic genetic genealogy as examining more than half a million SNPs, rather than the much smaller set of STR markers used in traditional forensic typing.

For a genealogist, the conceptual bridge is familiar. A consumer autosomal DNA test also surveys large numbers of SNPs so that shared segments can reveal relationships across multiple generations. The forensic difference is the source material. A consumer kit begins with a fresh saliva or cheek sample collected for testing. Forensic evidence may be old, chemically damaged, environmentally exposed, mixed with other DNA, or available only in a tiny amount. Laboratories therefore need methods designed for difficult evidence, documented handling, contamination controls, and a defensible chain of custody.

Othram describes its Forensic-Grade Genome Sequencing workflow as producing dense SNP data from forensic evidence and keeping laboratory processing, sequencing, and genealogy within a controlled forensic system. That description is one vendor’s model, not the only possible model, but it illustrates the stricter conditions around forensic samples: evidence preservation, accredited testing, data security, and an explicit decision about whether the expected investigative value justifies using material that may not be replaceable.

The result of Step 1 is not a suspect’s name. It is a dense genetic profile suitable for estimating biological relationships.

Step 2: Search only databases that permit the comparison

Once a usable SNP profile exists, it may be searched against genetic-genealogy databases that allow forensic or law-enforcement matching under their current rules. This is not the same as searching every consumer-testing company. Ancestry, 23andMe, and MyHeritage do not provide an ordinary law-enforcement matching lane through their consumer databases. GEDmatch and FamilyTreeDNA permit qualifying investigative matching under user-consent systems and platform-specific restrictions.

Consent is therefore part of the method, not an afterthought. A database match exists because a person tested, uploaded, or transferred data and because the kit’s settings make it eligible for that kind of comparison. The exact permissions differ by platform and sometimes by case type. GEDmatch, for example, distinguishes among privacy settings and between searches intended to identify perpetrators of qualifying violent crimes and searches involving unidentified human remains. FamilyTreeDNA provides a separate Investigative Genetic Genealogy Matching setting for customers who choose to participate.

The search does not hand investigators a folder labeled with the unknown person’s identity. It returns genetic relatives: people who share measurable DNA with the forensic profile. The useful information may include the amount of shared DNA, segment data, estimated relationships, shared matches, and whatever genealogical information participating users have made available.

The matches are often distant. DNA Doe Project explains that many useful matches are distant cousins whose common ancestors may fall in the 1800s or earlier. Some have well-developed trees. Others have no tree, a private tree, an incorrect tree, an adoption, an unknown parent, or a line obscured by endogamy or pedigree collapse. A strong match can shorten the search, but a case may also be solved by combining several modest matches that point toward the same ancestral couple or community.

For a hobby genealogist, this resembles an unknown-parentage problem. The stricter forensic rule is that the match itself must not be treated as evidence that the matching customer, or any named relative, committed a crime. The customer may be separated from the unknown contributor by several generations and multiple collateral lines. The match is a direction of travel.

Step 3: Build, document, and test family trees

This is the step most recognizable to an experienced genealogist. Researchers study the match list, cluster related matches, identify common ancestors, and build descendant trees forward toward the present. They work across census records, vital records, obituaries, newspapers, probate files, cemetery records, directories, yearbooks, public family trees, and other lawful sources.

The genetic evidence and the documentary evidence must be made to agree. A predicted relationship is a range, not a command. The same shared-centimorgan amount may fit several relationships. Half relationships, double relationships, endogamy, misattributed parentage, and pedigree collapse can all distort an apparently simple path. Researchers therefore test multiple hypotheses rather than attaching the unknown profile to the first plausible branch.

Good forensic genealogy follows the same habits serious hobbyists value: cite the record, distinguish an original source from a derivative claim, mark uncertainty, avoid copying unsourced trees, and resolve conflicting evidence. It must also add controls that private research may not require. Researchers should keep an audit trail of why a person was included or excluded, minimize unnecessary exposure of living people, separate verified facts from inference, and avoid contacting relatives in ways that could compromise an investigation or cause preventable harm.

Othram’s public description of tree integration emphasizes that matches alone do not resolve identity. Researchers map how matches relate to one another, locate common ancestors, and trace lines of descent until the unknown profile can be placed within a narrower family structure. DNA Doe Project similarly describes its work as a multidimensional puzzle in which even useful matches may lack public trees and require independent reconstruction.

Geography and chronology then do important work. A family may fit genetically but not historically. A candidate may have been too young, deceased, living in another country, or otherwise inconsistent with the known facts. Conversely, migration patterns, maternal surnames, military service, institutional records, or a cluster of relatives in the relevant location may strengthen one branch. These facts narrow the field; they still do not prove that a living candidate is the source of the forensic DNA.

Step 4: Turn the genealogy into an investigative lead — not a conclusion

At some point, the combined DNA and documentary work may narrow the search to one person, one sibling group, or one small family branch. This is the moment most likely to be overstated in public retellings. The genealogy has not “matched the killer.” It has generated an investigative lead.

That distinction protects both accuracy and the presumption of innocence. A family-tree hypothesis can be wrong because of an undocumented adoption, a mistaken parent, a sample problem, a misunderstood relationship, a record error, or simple researcher error. Even a correctly identified family may contain several people who fit the genetic distance. Investigators must therefore treat the name as a candidate to be checked, not as a verdict.

The Department of Justice policy repeatedly frames forensic genetic genealogy as a lead-generation technique. It also calls for collaboration among investigators, laboratory personnel, and prosecutors because the method affects privacy, evidence consumption, and later legal decisions. In practice, investigators compare the lead against facts that were not used to build the tree: age, sex, location, travel, access, known associates, physical description, case chronology, and other conventional evidence.

This stage should also work in both directions. A lead may focus an investigation, but it may also exclude a person or show that an earlier theory was wrong. Responsible use is not about making the genealogy fit the case. It is about testing whether the case facts and the independently built genetic hypothesis converge.

In an unidentified-remains investigation, the working lead may be a possible identity rather than a possible suspect. The same caution applies. A family connection and a plausible biography can justify targeted confirmation, but the person’s name should not be restored publicly until the responsible authority has completed the required identification process.

Step 5: Confirm through conventional forensic and investigative methods

The final step is deliberately outside the genealogy tree. Investigators seek independent confirmation.

In a criminal case, that may mean obtaining a lawful reference sample from the person identified by the lead and comparing it directly with the crime-scene evidence using validated forensic methods. Depending on the circumstances and jurisdiction, a reference sample may be obtained voluntarily, by court-authorized collection, or from an item lawfully collected during the investigation. The legal basis and laboratory method matter. The genealogy explains why investigators looked at a person; the direct forensic comparison addresses whether that person is the source of the evidentiary DNA.

In an unidentified-remains case, confirmation may involve DNA from a close biological relative, dental or medical records, fingerprints, radiographs, anthropology, personal effects, or a combination of methods accepted by the medical examiner, coroner, or other identifying authority. Othram’s discussion of reference testing notes that targeted relatives can confirm or refute a proposed kinship and resolve ambiguity among possible branches. DNA Doe Project’s cases likewise end with the relevant agency, not the volunteer tree-builder, making or announcing the official identification.

This separation is essential in court. A genealogical database search is not a substitute for a properly interpreted forensic comparison. Nor does a DNA source conclusion automatically answer every legal question. It may establish that a person contributed biological material, but context, timing, transfer, intent, and the remaining elements of an offense must still be proved under ordinary rules of evidence. The defendant remains presumed innocent unless and until guilt is proved in court beyond a reasonable doubt.

For hobby genealogists, the discipline can be stated simply: the tree proposes; the evidence disposes. Forensic genetic genealogy is powerful because it can turn distant cousin matches into a focused, testable hypothesis. It is trustworthy only when the people using it preserve the boundary between a promising lead and proof.

The method in one sentence

A forensic laboratory creates a dense SNP profile, searches only databases whose rules permit the comparison, genealogists build and test family trees from consenting matches, investigators turn the best-supported hypothesis into a lead, and conventional forensic work must independently confirm or reject it.

Keep reading: Database Permission Comparison — which DNA databases allow law-enforcement matching · How Forensic Genetic Genealogy Works · How DNA Remembers · the glossary.