Texas Man Close to Exoneration after Computer Algorithm Leads to New Suspect

Ellie Dessart

Mr. Ippolito

Forensics C Odd

23 March 2020

Current Event 19

Citation:

Ortiz, Erik. “Texas Man Close to Exoneration after Computer Algorithm Leads to

New Suspect.” NBCNews.com, NBCUniversal News Group, 17 Feb. 2020,

www.nbcnews.com/news/us-news/texas-man-close-exoneration-after-

computer-algorithm-leads-new-suspect-n1124946. 

            The article delves into the wrongful conviction of Lydell Grant in the 2010 stabbing of Aaron Scheerhoorn. The main evidence the prosecution had held against Grant was eyewitness testimony, along with Grant's history of “aggravated robbery, marijuana use, and theft.” Police had recovered DNA evidence, which consisted of a mixture of Scheerhoorn’s DNA and an unknown male, from fingernail scrapings on Scheerhoorn’s hand. At the time, Houston’s crime lab couldn’t conclude the genetic material was Grant’s. Further, Grant had an alibi the night of the crime. However, despite these holes in the prosecutor’s case, Grant was convicted of first-degree murder. In 2013, he reached out to the Innocence Project of Texas, and in 2018, his case was eventually referred to the Texas A&M School of Law. Students worked with DNA expert Angie Ambers, who ran the evidence through a software program called TrueAllele. The results confirmed that “Grant’s DNA did not match that of the unknown male profile.” After partnering with a crime lab in South Carolina, which had access to the FBI database known as the Combined DNA Index System (CODIS), the Innocence Project came across a DNA profile that matched the genetic material from the fingernail scraping— it belonged to a man named Jermarico Carter, who later confessed to the killing. The end of the article describes the controversy over the accuracy of systems like TrueAllele. Unfortunately, the source code for many of these softwares are “trade secret” since they “must be protected in a highly competitive commercial environment,” so the question of whether these programs can be trusted is up for debate. 

            The pending exoneration of Lydell Grant highlights the impact of DNA technological advancements. When the Houston crime lab first examined the evidence, they couldn’t determine with certainty whether or not Grant matched the unknown profile, because at the time, it was difficult to manually separate and interpret a mixture of DNA from more than one person. Almost a decade later, we now have access to software and databases that facilitate the process and do what scientists years ago couldn’t do by hand, which is remarkable. However, with these developments come drawbacks. Since companies such as TrueAllele don’t release their source code, it’s difficult to determine how accurate the systems truly are. While the development of such software can help solve cold cases and bring justice to wrongly convicted people, it could also, if inaccurate, send innocent people to prison. As the forensic world continues to adapt and evolve, it’s important to not only celebrate our progress, but to also be aware of the shortcomings that still exist. 

            Overall, I found the article compelling and crucial to our study of forensic science and the law. It’s upsetting to see how a man lost years of his life after being convicted on the basis of very little proof. Instead of focusing on scientific evidence, the jury decided to trust the eyewitness testimony, which, as we know, isn’t always reliable. The author, Erik Oritz, organized the piece well by first discussing Grant’s case, then describing the journey to free him, and finally ending with the controversy over certain genetic databases. The article was thorough and did a good job exploring both the pros and the cons of this emerging form of DNA technology. However, the article was lengthy and dense. I had trouble following along with Grant’s case in the first half and the debate over genetic software in the second. Additionally, I struggled to keep my own summary concise because there was so much to write about. To improve the article, I’d consider splitting the piece into two separate stories, published as Part One (Grant’s case and the journey to free him) and Part Two (the question of whether these genetic systems can be trusted). Shorter articles would help keep readers engaged and interested in the story.