17-Feb-2017

Cybergenetics scientist presents three AAFS talks

New Orleans, LA

On Thursday, February 16, Professor Fran Watson (Indiana University McKinney Law School, Wrongful Conviction Clinic) and Cybergenetics Dr. Mark Perlin spoke about the TrueAllele® exonerations of Darryl Pinkins and Roosevelt Glenn. Exculpatory DNA evidence data was available over fifteen years ago. But the common failure to fully interpret DNA data cost Pinkins fifteen extra years in prison.

On Friday, February 17, defense attorney Clinton Hughes (Legal Aid Society of New York City) and Cybergenetics Dr. Mark Perlin spoke about the dangers of using subjective DNA methods beyond their tested limits. When analysts are "forced to pick and choose data" and "input parameters" for their software, and lack validation guidance, their scientific results may be unreliable.

That afternoon, Cybergenetics Dr. Mark Perlin spoke about customizing error rate to DNA evidence. A likelihood ratio (LR) summarizes evidence in a match statistic. Also important is the LR error - the chance of falsely identifying an innocent person at that LR level. His talk introduced false match probability (FMP) error, and showed how to rapidly calculate FMP.

Links

• When DNA alone is not enough: exoneration by computer interpretation - Cybergenetics
• On the threshold of injustice: manipulating DNA evidence - Cybergenetics
• Error in the likelihood ratio: false match probability - Cybergenetics