## Facts

- Frank Lee Soto was charged with forcible rape of a 78-year-old woman and with knife use; the assailant wore a mask, preventing visual identification.
- Police recovered a bedspread with semen stains from the victim’s bedroom and obtained Soto’s blood sample for comparison.
- A criminologist performed RFLP DNA testing, compared autoradiographs from the semen and blood samples, and concluded the relevant DNA fragments matched.
- The expert calculated a random match probability using an Orange County “Hispanic” population database and the unmodified product rule, testifying the probability of a random match was about 1 in 189 million.
- The jury acquitted Soto of forcible rape and knife-use allegations but convicted him of attempted rape as a lesser included offense.

## Issues

1. Under California’s Kelly standard for novel scientific evidence, are DNA random-match probabilities calculated with the unmodified product rule admissible in criminal trials?
2. Is the unmodified product rule generally accepted as reliable in the relevant scientific community when applied to population frequency data in forensic DNA analysis?

## Decision

- The Supreme Court of California affirmed the judgment, in a unanimous opinion.
- The court held that random-match probabilities calculated using the unmodified product rule are generally accepted in the relevant scientific community and satisfy Kelly.
- The trial court did not err in admitting the DNA statistical evidence, and the attempted-rape conviction stood.

## Legal Principles

- Under Kelly, novel scientific evidence is admissible if: (1) the method is generally accepted as reliable in the relevant scientific community, (2) the testifying witness is qualified, and (3) correct scientific procedures were followed in the case.
- Forensic DNA match evidence may be accompanied by quantitative random-match statistics when the statistical method meets Kelly’s general-acceptance requirement.
- General acceptance does not require unanimity; methodological criticisms and debates about assumptions (e.g., database representativeness, population substructure, independence across loci) ordinarily affect weight, not admissibility, and are addressed through cross-examination and competing experts.

## Conclusion

The court approved admission of DNA random-match probability testimony calculated with the unmodified product rule under the Kelly standard, treating disputes over databases and statistical assumptions as matters for the jury’s evaluation rather than categorical exclusion.


People v. Soto, 21 Cal. 4th 512, 981 P.2d 958 (Cal. 1999)