to hundreds of thousands of markers, and with that information, we can detect more—dista nt relationships. every time you test for dna, you lose a bit of the sample in the process. in carla's case there was little of the killer's dna left. it was likely this was the last roll of the dice. we had to ask ourselves, "have we seen enough dna that is of this kind of quality and property, to where we feel confident that there's a good chance we'll have a positive outcome?" otherwise, we don't want to do it, cos when you test dna, you're consuming it, so you're destroying evidence. once they'd sequenced the killer's dna, they ran it through several genealogy websites. from there, they created a family tree of the killer. and then began to look for possible suspects — the right age, male, who lived in texas at the time of the murder. they developed a theory, that the dna found at the scene belonged to a man called glen mccurley, a man who lived close to carla, a man who had previously