Is DNA evidence more convincing than fingerprints?

Innocent Behind Bars (1/2)Wrong tracks

Charles Fain: "I don't trust scientists too much."

John Butler: "You have to make a decision. Is this real or not?"

Greg Hampikian: "Innocent people rot in prison."

John Butler: "Guilty or Not Guilty?"

Charles Fain: "A lot of science out there is junk."


When nine-year-old Daralyn Johnson disappeared in the town of Nampa, Idaho, Charles Fain was sleeping in his father's trailer - in Redmond, Oregon, less than half a mile away. It was Wednesday, February 24, 1984. Charles Fain: "I had no idea that a year and a half later I would have to prove where I was."

Charles Fain was 32 at the time. After returning from the Vietnam War, he had alcohol problems. "I went to jail a couple of times. So I had a criminal record. I stopped drinking on April 1, 1979. And never started again." He found work sometimes for six months, sometimes for a few days. Sometimes in Idaho, sometimes in the neighboring state of Oregon.

For a year and a half, the Nampa police groped in the dark looking for Daralyn's killer. On the day of the crime, a girl had seen a car from the 60s, early 70s. Charles Fain: "They went through the approvals and pulled out a bunch of cars and mine was one of them. But they couldn't find me, I was in Oregon." When Charles Fain returned, the police called him in.

In the end, the wrong person is always in jail

You know that from CSI. Murder, police. The forensics department powdered everything with black powder, pours plaster of paris in footsteps, and collects hair and textile fibers from the corpse. After more or less twists and turns, the right person is in jail.

The reality is of course different. Although some of the methods have been in use for well over 100 years, they are much less reliable than generally assumed. And in the end the wrong person is always in jail. Several high-ranking scientific commissions in the USA have come to this conclusion in recent years, most recently the PCAST. John Butler: "The United States President's Council of Advisors on Science and Technology presented its report in September 2016."

President Barack Obama had commissioned this scientific advisory body to clarify what is behind the fact that up to 100,000 people in the USA are imprisoned for crimes they have never committed. The PCAST ​​writes about these statistics in its report:

"Independent reviews [...] have shown that many cases were based in part on incorrect expert statements from forensic experts. They had told the juries that samples from the crime scene (for example hair, bullets, bite marks, tire or footprints) were highly likely to be accused associated with a crime. "

High probability. This is rarely heard in the relevant television series. In doing so, it is essential to know how reliable a trace is - or just how big the doubt is about it. And to quantify that, you need a decent database. But most forensic methods lack that, says Dr. John Butler, senior forensics expert at the United States National Institute for Standards and Technology in Gaithersburg, Maryland.

"How often does a certain characteristic occur, for example a random change in shoe soles? That is not known. There must now be studies so that we have figures that really say something. So far, the experts have only made qualitative statements: 'In the mirror in my experience it looks like the shoe. ' But how do you come to this experience? What does the database look like? That is ultimately the core of the PCAST ​​report: You need data to support your conclusions and assertions. "

"FBI and other agencies have done a lot of nonsense"

Professor Peter Schneider heads forensic molecular genetics at the Institute for Forensic Medicine at the University Clinic in Cologne, and he knows what is behind the criticism. Hair comparisons are particularly bad.

"In fact, there was a lot of nonsense done by the FBI and other agencies in the US in the 1990s and early 2000s, and unfortunately to the point where people were punished with death."

The deficiencies are part of a structural problem: The Justice Department and the Federal Police FBI have checked more than 3,000 criminal cases in which hair comparisons played a role. In 95 percent of the cases, the hair experts of the FBI would have made invalid statements from a scientific point of view.

Like Charles Fain. “They asked me about hair and I gave them some. I never thought I'd hear from them again. Five or six weeks later they checked me in and said, 'The hair proves it was you.' I said, 'This is impossible!' "
But the police did not give up. Charles Fain: "They had a case that they couldn't solve. And suddenly they had a suspect. I was their only hope."

"Project for the Innocent" proves the innocence of 13 people

Greg Hampikian comes down the hall on the second floor of the brick building that houses Boise State University. Large dark eyes under short white-gray hair, a gray three-day beard around the full lips. His office is tiny. Two computers share the desk space and a colorful model of the DNA double helix. Genetic material is Hampikian's area of ​​expertise.

At the beginning of his career he researched the Y chromosome. That was when he got into forensics. "Then in 1999 I heard from Calvin Johnson, who had been freed by DNA and the Innocence Project. He was jailed for 17 years for rape."

Greg Hampikian heads the Idaho Innocence Project and fights against systematic errors in forensics (Deutschlandradio / Joachim Budde)

Greg Hampikian founded an Innocence Project for the state of Georgia with students from Emory University in Atlanta. Shortly after the turn of the millennium, he moved from the southeastern USA to the northwest, to the Rocky Mountains in Idaho.

"I am a professor at the Biology Faculty and also give lectures on forensics at the Chair of Criminal Law. In our work in the Innocence Project, students from all departments and across the country help us."
The Idaho Innocence Project is the only one that has its own DNA laboratory. Greg Hampikian studies ancient cases from around the world. To date, he and his helpers have proven the innocence of 13 people. At the same time, he is researching the weaknesses of forensics and looking for ways to improve them.

Wrong tracks, wrong testimony

Of course, it takes more than a few hairs to pin someone as a culprit. With Charles Fain, a vague description of the perpetrator fit: a man with medium-length brown hair and a beard. The investigators also found shoes that they attributed to prints at the scene.

Charles Fain: "There was only one problem: I had only bought the shoes six months after the crime, but I no longer had a receipt to prove it. I almost threw them away. They were worn out anyway."

Charles Fain never confessed to the fact. But others claimed the opposite. Charles Fain: "I told a friend about the hair test. And later he told the cops that I had confessed to him on the occasion. A really nice friend."

It got worse: while in custody he was in a cell with other prisoners. Charles Fain suspects the police used her as an informant and promised them a deal. Two claimed that he had also admitted the crime to them. They made convincing appearances in court. Charles Fain sat in the courtroom with shackles and watched in disbelief. "I didn't know what to do. I just sat there thinking, 'Oh, my lawyer knows what he's doing.' But it was too much for him. "

On November 4, 1983, the jurors found him guilty. On February 17, 1984, the court determined the sentence: death sentence. Charles Fain: "I was numb. I wasn't really thinking anything. I went back to the cell, sat on my bunk and said, 'I can't believe it.' It's really difficult to put into words how you feel. "

"DNA analysis is one of the most powerful pieces of evidence"

Greg Hampikian: "DNA is very special."

The investigative authorities have been using genetic fingerprints since the mid-1980s. If the police find material that contains body cells at the crime scene, geneticists can create a DNA profile based on distinctive features in the genome. If they have a suspect, they do the same and can compare the profiles. This principle has not changed to this day, says Peter Schneider.

"The DNA analysis in criminal proceedings is still one of the most powerful and most secure pieces of evidence that we know. As a result, it has also revolutionized the way perpetrators are investigated and traces are processed.

Initially, the DNA fingerprint was considered infallible - a nimbus that continues to this day. But as early as 1989 a court in New York declared a DNA trace inadmissible because a reliable database was lacking. Geneticists then described which locations in the genome are suitable for analysis and which variants - called alleles - occur in humans and how often.

"The gold standard is samples from a single source." NIST's John Butler is one of the world's leading experts in forensic DNA analysis. He explains the different degrees of difficulty in DNA analysis with an analogy from mathematics:

"If a DNA sample comes from just one person, that's simple arithmetic: two plus two equals four. In a sexual offense, when we have the victim and a perpetrator, that's algebra. We know the victim, so we have one Equation that we have to solve after the perpetrator as an unknown. So far no problem. But if we get traces of contact where three or more people have touched a surface, then that's complicated stuff. "

Complex DNA mixtures cannot identify a single person

An equation with several unknowns. That's higher math. Greg Hampikian: "Such complex DNA mixtures are not like traditional DNA fingerprints. They cannot identify a single person. In fact, most of the people who cannot be ruled out did nothing to the mixture. This is not a science like a pregnancy test. "

Each gene consists of base pairs, represented by the letters A and T as well as G and C. In addition, many base pairs in the genome do not belong to the actual genes. There are short sequences of two to six of these base pairs between the genes. These sequences have no known function and they repeat themselves. The experts speak of microsatellites or short tandem repeats. These repetitions are counted by the geneticists for the DNA fingerprint, says John Butler.

"They don't code for hair color, eye color or size. We just measure the length: sometimes a 12, sometimes a 14, sometimes a 15. This avoids the risk of violating someone's privacy."

Geneticists use 13 to 20 such locations in the genome for a DNA profile - depending on the country. At each point, researchers have found a number of variations in the population - called alleles. Once all of the gene locations have been counted from a DNA trace, forensic experts draw a curve from which peaks protrude for each allele. The likelihood that two people will have the exact same profile is more than one in a hundred trillion - at least in theory.

The problem arises with mixed trace elements: the individual alleles have no name tag. Greg Hampikian of the Idaho Innocence Project tried to illustrate this to a prosecutor with a bowl of Scrabble letters:

Greg Hampikian: "I told him: Let's take the letters of both of our names. They were A, E, G, H, I, K, L, M, N, O, P, R and Ypsilon." Then he made a list of names that can be put together from these letters. "Al Gore and all of these others might as well be misread. Just because someone can't be disfellowshipped doesn't mean they actually contributed DNA. It's difficult to get that to a jury.

Contamination can lead to incorrect results

On March 24, 2009, two women were raped in a warehouse. The police use the semen on their underwear to determine the alleles of the Y chromosome, the chromosome that only men carry. Everything points to the three men who rented the warehouse.

Greg Hampikian: "All of the alleles on your Y chromosomes are in the seed spot. Sounds like a clear case."

You will be convicted based on the DNA profile. One, Chen Long-Qi, protests his innocence. But he too has to go to jail because the DNA expert says he cannot exclude him.

Greg Hampikian: "The Taiwan Association for Innocence, which I worked with on this case, found that there is a new test that looked at six new markers. You can see that Mr. Chen is excluded in two places. "

Peter Schneider: "We have become the victim of our own success." New tests can get innocent people out of jail. The new precision can also get people into trouble in the first place. Greg Hampikian: "Most laboratories can create a very decent DNA profile from 20 sperm cells."

The "angel with the eyes of ice", Amanda Knox was innocent in prison for four years. (dpa / picture alliance / Pietro Crocchioni)
Amanda Knox was the "angel with the eyes of ice" in the media. Perugia, November 1st, 2007: The British student Meredith Kercher is brutally mistreated and murdered. The police find Rudy Guede's finger and hand prints and his DNA in the victim's vagina. He confesses and incriminates Meredith Kercher's roommate Amanda Knox and her friend Raffaele Sollecito. The police found the alleged evidence in Sollecito's kitchen: a knife with no blood on the blade, but the victim's DNA. Amanda Knox and Raffaele Sollecito are found guilty.

Greg Hampikian: "This sample of the blade matched the alleles of the victim who never entered Sollecito's apartment. At first glance, it looks like a great piece of evidence."

Greg Hampikian is a defense expert in the audit department. The defenders are challenging the DNA evidence. Greg Hampikian shows the analysis profile. A string of tips. The special thing is: "There are no peaks above 113. Most are even below 100, some even only 30, 27, 21. At that time, such pricks would have been considered invalid in most laboratories. Today there are laboratories that use this method reliably mastered, but only many years later. "

The defense argues, and later studies confirm: These tiny spikes were from impurities. The police transferred the victim's genetic material to the knife while handling it. The highest Italian court acquitted Knox and Sollecito. They were innocent in prison for four years.

Study: Most of the experts come to the wrong conclusion

The difficulty can now be determined: Which DNA from the crime scene really has something to do with the crime? Something else is added:
John Butler: "We have the problem: If we send a mixed sample to two different laboratories, we can get two different results back. This is not new, it has been known for 15 years or more."

John Butler and colleagues from NIST last carried out a study in 2013, in which they sent data from trace mixtures to 108 DNA laboratories in the United States. Only seven laboratories found the correct solution. 76 laboratories assigned the data to someone who had not contributed to the mixture. 25 came to the conclusion that the data situation was inconclusive.

Greg Hampikian of the Idaho Innocence Project criticizes the fact that Butler and his colleagues present the NIST results at specialist conferences, but do not publish them in a science magazine. Or in one of Butler's textbooks.

Greg Hampikian: "NIST has shown that the majority of the experts in their study come to the wrong conclusion. This is important information! If it were about seat belts or airbags or peanut butter, there would be a nationwide investigation and everyone would be on their hind legs But because it's about inmates, I think the excitement is less. "

And what about in Germany? Peter Schneider from Forensic Molecular Genetics in Cologne assesses the situation as follows: "I do believe that we have a relatively good quality here in Germany due to our round-robin tests in the field of professional laboratories that have been running for over 20 years, because we meet once a year put them together and always discuss our result. "

18 years innocent on death row

Charles Fain, the convicted child murderer and rapist from Idaho, has always protested his innocence. It took him many years, a new lawyer, and several attempts before a judge would re-examine his case. In 2001, a judge allowed the hair investigators found on the body to be examined for DNA. It was now possible.

Charles Fain: "You first examined a hair that didn't belong to me.And then they tested all the hair they had: scalp hair, pubic hair, armpit hair. They also excluded me. "The judge offered him a deal." 52 days later they let me out. "After 18 years on death row, the real culprit was never found.

The evaluation methods are slowly catching up with the analysis methods. Computer programs calculate how likely it is that someone contributed DNA to a mixture. This method is called Probabilistic Genotyping.

Greg Hampikian: "Probabilistic Genotyping tells us which peaks in the analysis curve or which Scrabble letters belong together. Which names are hidden in the mixture? Which DNA profiles?" At the end of 2016, Greg Hampikian, the molecular biologist and head of the Idaho Innocence Project, was able to use this method to prove the innocence of two people convicted of rape.

Hammond, Indiana, effectively a suburb of Chicago. On December 7, 1989, a 27-year-old woman was waiting in her car at a red light when a car hit her from behind. The woman gets out to examine the damage. Several men in green overalls come out of the other car. One of them pulls her into the car, they kidnap and rape her.

The police find the overalls and associate them with workwear from a company where Darryl Pinkins and Roosevelt Glenn work. A colleague also owns a 1973 Pontiac, as described by the victim. But that's not all: the woman identifies Darryl Pinkins as the man who is said to have dragged her into the car. He has an alibi. Unfortunately also the right blood group.

Greg Hampikian: "When the lab determined the blood types, they said, 'We can't rule out the suspects.' But blood type tests are not very statistically meaningful. I think about 70 percent of African Americans could not have been excluded from this mix of blood types. And 100 percent of Europeans. Such a thing is meaningless as evidence. Nevertheless, the prosecution used it. "

In one study, only seven out of 108 DNA laboratories in the US came up with the correct solution. (Sven Hoppe / dpa)
In January 1990 Darryl Pinkins, Roosevelt Glenn and three other men are charged. Greg Hampikian: "Before the trial, the prosecution had the traces tested for DNA, which was brand new at the time. The lab identified two main culprits, and all of the defendants were excluded. So they had the wrong ones. But the prosecution insisted: 'They match the blood types.' And that put these poor men behind bars. "

From 2006 onwards, Innocence projects have been working on this case. With Probabilistic Genotyping they had a new method and thus a reason for the revision. Greg Hampikian: "This new technique helps us to analyze the DNA data. The data with which I do a bunch of calculations like with a slide rule, from which a computer can analyze 100,000 versions, change the variables and in the end determine:" 80 percent of the time, we get DNA profiles that look like this. "So it's a probability-based analysis of the profiles.

Darryl Pinkins will be released on April 22, 2016, after more than 25 years. He is the first to be proven innocent by Probabilistic Genotyping. He and Roosevelt Glenn are now fighting for their full rehabilitation.

Greg Hampikian received $ 630,000 from a US Justice Department grant in May. He estimates that with the money he will be able to investigate twelve cases using Probabilistic Genotyping. Hundreds of prisoners pleading innocence are available to choose from.

The genetic fingerprint is also being further developed

The PCAST ​​report requires that other traces are also placed on a proper database. John Butler, NIST's DNA expert, sees a lot of work to be done in forensic science. "There are a few such studies, some are still missing. And where they exist, they have not yet been published."

Peter Schneider remains skeptical. "That is and remains a pattern comparison. And you will never be able to get to the point where you can actually produce a reliable statistical quantification."

The genetic fingerprint is also being further developed: instead of 13 gene locations, US laboratories have been using 20 markers since January 1, 2017. This has three advantages: First, the database has now grown to 16 million DNA profiles. Just as a large city needs more digits for the phone numbers, the database also needs more markers to distinguish the entries quickly and reliably.

Second, the expansion helps with international searches: because the new markers are also used in Europe. Last but not least, the new markers make it easier to work with DNA that is already damaged. Greg Hampikian, the forensics professor: "I don't know what techniques we'll have in ten years. I don't know if anything will be superior to DNA. I don't know if we'll be able to tell how DNA got somewhere, so whether it leaves traces when it is moved through space and time. If that succeeds, we could answer a lot of questions about transfer and contamination, i.e. how DNA got where. "

As early as 2009, the United States National Academy of Sciences pointed to forensic issues. In 2013, the Ministry of Justice established the National Commission on Forensic Science. By January 2017, she had developed a series of advice and information on how to improve the quality of forensic work. The Trump administration's new attorney general, Jeff Sessions, has stopped the reform efforts that were launched under Barack Obama. He said, according to the Washington Post newspaper:

"We shouldn't imply that proven scientific principles that we've used for decades are in any way unsafe. Otherwise, law enforcement officers will have to fend off attacks on the most fundamental issues in court."

Good science - bad science

Charles Fain was on death row for 18 years. To date, he has not received a cent in compensation. "I'm not angry with the system."
John Butler. "Any technology can be improved. That's what science is about."

Charles Fain. "I just decided: I let God work in my life. The whole story is not worth getting upset about. You have to do it, otherwise it is they who control you."

Greg Hampikian. "But you shouldn't take away the teaching that DNA is bad, it isn't. DNA is great. It's like McDonalds. There are people involved. Sometimes they forget your soda."

Charles Fain: "Bad science got me behind bars. Good science got me out again."

The Dlf broadcasts the second part of the focus on August 6th, 2017 at 4.30 p.m .: Innocent behind bars. On Errors in Forensics, Part 2: Fallacies