Genetic genealogy turns a 1993 homicide into a solvable case
For decades, Randy Gail Sperino’s killing sat in the category that haunts detectives and families alike: a violent death with evidence, but no name attached to the person responsible. The 34-year-old was last seen alive on 9 November 1993 in Granite City, Illinois, and her body was found hours later in a nearby field after blunt force trauma. Standard forensic DNA testing did not produce a match in law enforcement databases, leaving investigators with a crime scene profile and no direct path to a suspect.
A fresh DNA strategy changed that trajectory. Investigators in Madison County sent genetic evidence to Othram, a Houston-based company that uses investigative genetic genealogy to identify suspects and unknown remains in cold cases. Unlike traditional forensic DNA testing, which depends on a direct comparison with a known person already in a database, this approach searches for distant relatives of an unknown individual and then works outward through family trees. In Sperino’s case, the killer’s own DNA profile was not available, but profiles submitted by extended relatives helped identify a male family member.
That single lead reopened a case that had already consumed 30 years of interviews, dead ends, and investigative labor. It led detectives to Albert “Buddy” Zigler, now 70, who later confessed and was charged in May. For Sperino’s family, the scientific breakthrough carried the kind of finality that ordinary database searching could not deliver. Wes Sperino said at a press conference, “I finally get closure and I finally know who did this.”
The case reflects a shift in forensic practice that is still gaining ground. Investigative genetic genealogy proved its value in the 2018 identification of the Golden State Killer, Joseph James DeAngelo, and the same logic now extends to unsolved homicides and unidentified human remains. The method is powerful because it does not require a direct suspect match; it exploits the genetic breadcrumbs left by relatives who have opted to share DNA data with consumer ancestry platforms or similar databases used with consent for law enforcement matching.
Why distant relatives can expose an unknown suspect
The biological premise is straightforward even when the investigative work is not. People share segments of DNA with relatives, and those shared markers can point to a family network even when the perpetrator has never submitted a sample. Othram’s David Mittelman compared the process to GPS: the DNA profile provides the signal, but scientists still need a map to interpret where that signal belongs. Public records, obituaries, census data, and family-history research fill in that map until a likely branch of a family tree comes into focus.
This is where forensic science shifts from a laboratory exercise to a broader investigative process. DNA may narrow the field, but it does not close the case on its own. Mittelman stressed that the work often stops being about genomics and becomes a sequence of interviews, record checks, and old-fashioned police work. The technology can suggest where to look; detectives still have to walk the final miles.
That limitation is built into the method. If there is no usable DNA, investigative genetic genealogy offers no route forward. When a profile exists, though, it can be far more productive than waiting for a conventional database hit. In practice, that makes it especially useful in cases where biological evidence was preserved long enough for modern methods, but not well enough to identify the perpetrator through standard comparison alone.
AI helped investigators make sense of 2,000 pages of records
DNA was not the only tool driving the arrest. Madison County detectives also used artificial intelligence to manage the paper trail accumulated since 1993. The team fed more than 2,000 pages of records into Centerline AI, a platform from Florida software company CentralSquare, to help reconstruct timelines, organize witness information, and identify evidence worth testing. For an investigation stretched across three decades, that kind of document analysis can change the pace of the work.
One sentence captures the practical impact: the AI made it easier to process years of records without asking a detective to read every page manually. Lawrence said his team could interact with the system like a chatbot, using it to surface connections and organize material that might otherwise have remained buried. The platform also helped during the interview phase, when detectives used it to examine whether Zigler’s statements were truthful and to guide follow-up questioning.
The technology did not replace investigative judgment. Lawrence was clear that there comes a point when software has to give way to street-level police work, witness interviews, and direct confrontation with facts. AI can sort, summarize, and reassemble large quantities of evidence. It cannot decide credibility, persuade a suspect to confess, or substitute for the human judgment needed when a timeline still contains gaps.
Privacy, funding, and the limits of access
The spread of investigative genetic genealogy raises privacy concerns that cannot be brushed aside. Denise Syndercombe Court, professor of forensic genetics at King’s College London, pointed to the central ethical tension: one person may consent to law-enforcement matching through a consumer database, but genetic relatives may never have agreed to have their DNA indirectly implicated in an investigation. Family relationships make consent difficult to contain within the individual who uploads a sample.
Funding creates a second barrier. Othram says it has helped solve thousands of cases, but Mittelman said many agencies still cannot afford to use the approach routinely. Some cases rely on donations or crowdfunding, and Othram maintains a website where members of the public can help finance selected investigations. The spread of the technology therefore depends not only on scientific validity, but on budgets, access, and whether agencies can support the downstream work required after a genetic lead emerges.
The UK is evaluating whether the method should be adopted more broadly, though it is not yet in use there. Syndercombe Court argued that Britain already has an efficient DNA database and can identify a person from crime-scene DNA in more than 65% of cases when usable DNA is available. Even so, investigative genetic genealogy may matter most in the cases that conventional systems cannot solve: old evidence, no direct database hit, and a family trail waiting to be mapped.
The Sperino case shows how modern forensics now operates at the intersection of molecular biology, data science, and patient persistence from investigators. A DNA sample pointed to a family, AI organized decades of records, and detectives converted those leads into a confession and murder charge. The unresolved question is not whether the tools can work, but how often agencies can afford them, how they will be regulated, and how far consent should reach when one person’s DNA can reveal another person’s identity.