Imagine being able to create a real portrait of a person using a DNA sample. It sounds like something straight out of a Hollywood blockbuster, but new research in the Centre for Forensic Science is turning science fiction into reality. The implications for identifying victims of natural disasters and unsolved crimes are enormous.
During the last 25 years the use of DNA profiling has absolutely revolutionised forensic investigation. It has provided an unprecedented level of sensitivity, specificity and statistical significance.
The current use of human DNA for the purpose of individual identification relies strictly on comparative grounds – DNA profiles obtained from crime scene material are compared with those of known potential suspects or in the case of paternity, with the alleged father. Similarly, in mass disaster or missing person cases, DNA profiles obtained from an unknown person are compared with those of known relatives, or with direct reference samples from items belonging to the missing person.
During my career as a forensic DNA officer overseas, I saw a lot of serious criminal cases. In many of them, I was able to obtain a DNA profile – that unique ‘barcode’ used in human identification. From that, I was often able to find a matching suspect’s profile. Bingo – case solved! However, in others, where the suspect was not available or no match was produced in the DNA database search, the so-powerful DNA evidence provided no help in solving the crime.
Now, imagine if we could ‘convince’ the DNA molecule to ‘talk’ and give us information about the physical appearance of the person it came from. This kind of information would be absolutely invaluable.
In the last few years this rationale has given rise to a brand new forensic discipline: Forensic Molecular Phenotyping or in other words, ‘Investigative Genetics’. This field of research seeks to obtain additional information about the source of a DNA sample, which might generate an investigative lead such as skin, eye and hair pigmentation, ancestry and more recently, facial morphology.
And that is exactly the focus of my research. In my recent PhD project I was analysing thousands of specific ‘bits’ of our DNA that are responsible for the differences in the way we look. For example, our face – the size and shape of the nose, eyes, ears and other facial features like the eye lids and ear lobes, or our pigmentation (the colour of eyes, skin and hair). These specific ‘bits’ are called single nucleotide polymorphisms (SNPs) and represent a single change in our DNA.
There is a remarkable variety of human facial appearances, almost exclusively the result of genetic differences. However, the majority of the genes and specific genetic variants that affect the size and shape of the cranium and the soft facial tissues are still largely unknown.
In order to find the specific SNPs that influence our facial morphology, I collected DNA samples and 3D facial images, along with extensive phenotypic information and ancestry, from almost 600 volunteers. Those images were analysed with graphical software to collect over 100 craniofacial measurements. Then I used next-generation sequencing platform and bioinformatical software to genotype and analyse the data to see which SNP corresponds to which facial features and other externally visible characteristics.
Surprisingly, I was able to find significant associations between 19 craniofacial traits and 78 SNPs, many of them novel (a summary of these results has been submitted for publication).
The results will not only enhance our understanding of the genetics that underpin the normal craniofacial morphology, but will also be particularly valuable in the forensic field, allowing the prediction of a person’s appearance from a DNA sample.
How so? De-coding the DNA information about physical appearance will enable the development of a real image from a molecular identikit that could help to identify the offenders and assist with facial reconstruction in unidentified skeletal remains and disaster victim identification cases.
Eight months ago I was appointed as a new Postdoctoral Research Associate at UTS in Associate Professor Peter Gunn’s forensic molecular biology group. Soon I will begin collecting more samples so that I can continue and extend my previous research. I’m aiming to collect over 1000 DNA samples and 3D images and am looking for volunteers to participate and support this project.