"Inconclusive" three-person mixture yields guilty plea: TrueAllele genotype revival

Kelly Woolard, Dr. Mark W. Perlin, & Dr. Kevin Miller, "'Inconclusive' three-person mixture yields guilty plea: TrueAllele genotype revival", Promega's Twenty Fifth International Symposium on Human Identification, Phoenix, AZ, 1-Oct-2014.

Poster

Abstract

In May of 2011, a Kern County prison guard saw inmate Anthony Ford throw a shiv (i.e., homemade weapon) down onto the ground. The shiv was a screw attached to a handle, bound with elastic string. Mr. Ford was charged with possession of a manufactured weapon in prison.

In November of that year, the police sent the shiv to the Kern Regional Crime Laboratory (KRCL) for DNA analysis. In March of 2012, a KRCL analyst advised the Deputy District Attorney that touch items do not produce good DNA results, and are often inconclusive. So in May the unanalyzed shiv was returned to the police.

In November of that year, the police sent the shiv to the Kern Regional Crime Laboratory (KRCL) for DNA analysis. In March of 2012, a KRCL analyst advised the Deputy District Attorney that touch items do not produce good DNA results, and are often inconclusive. So in May the unanalyzed shiv was returned to the police.

The private lab calculated combined probability of inclusion (CPI) statistics. 1 of 4 Caucasians could not be excluded from the handle mixture, 1 of 8 African-American could not be excluded, and 1 of 6 Hispanic persons. The defense attorney elected to not use these relatively uninformative CPI results.

In 2013, the KRCL prepared to interpret DNA mixtures using its new TrueAllele^® Casework system. Their TrueAllele validation study examined up to five unknown contributors on laboratory-prepared mixed samples. Forensic analysts were trained and certified on how to use the probabilistic genotyping system. On October 10, 2013, the KRCL deployed TrueAllele in-house for automated computer interpretation of forensic DNA evidence. By year's end, the KRCL was poised for computer analysis of complex DNA mixtures.

On February 7, 2014, KRCL received the private lab's .fsa electronic data files, and entered them into TrueAllele. The computer's genotype modeling excluded Ford from the screw and the string. However, TrueAllele found that a match between a minor contributor to the handle and Anthony Ford was 1.4 million times more probable than coincidence, relative to a Caucasian population. The statistic was 2.4 million for African-Americans, and 1.6 million for Hispanics.

Following defense discovery on the KRCL's TrueAllele DNA match results, Mr. Ford pleaded guilty to the charges. Since TrueAllele separated the DNA mixture data into three genotypes (one for each unknown contributor), the computer had also inferred the shiv's major profile. Upon entry into CODIS, this TrueAllele-derived profile yielded an offender hit from the Cal DOJ DNA data bank.

In this touch DNA case, a government lab's TrueAllele computer analysis solved a low-level three-person mixture that human CPI analysis could not. The major contributor produced a CODIS hit, while a minor contributor provided inculpatory DNA evidence that led to a guilty plea.