FDNA has created a new type of algorithm which may be able to help diagnose rare genetic conditions like Noonan Syndrome by a simple picture of a person’s face.
What’s particularly cool about their algorithms is that they can determine not only whether or not someone has a genetic mutation but which specific gene is mutated, even within the same disease state.
This algorithm was examined with Noonan syndrome.
Noonan Syndrome
Noonan Syndrome does not look the same across patients. Mutations in a wide array of genes can cause the condition, and symptoms present themselves differently across individuals. These can include short stature, learning disabilities, heart complications and unique facial appearances. For some patients, the condition puts them at a much higher risk for developing leukemia.
Two of the most serious forms of Noonan syndrome are caused by mutations in the RAS1 gene and the KRAS gene. Mutation in the RAS1 gene frequently leads to hypertrophic cardiomyopathy. A mutation in the KRAS gene can result in severe neurological outcomes. Unfortunately, there are no treatments currently approved to treatment Noonan syndrome.
This is the case for too many rare conditions. However, without proper diagnosis, we can’t conduct the research we need to develop novel therapies.
The algorithm
Researchers conducted a study with these new algorithms explicitly focused on Noonan syndrome however they make it clear that this technology Is meant to help diagnose a wide array of disorders. In the study, the software was analyzed to determine its ability to uncover which of five different mutations caused a certain individuals Noonan diagnosis. The algorithm was correct for 64% of patients. If clinicians had predicted this diagnosis on their own (without using genetic testing), it is expected their success rate would only be 20%.
You can read the full study here, published in Nature Medicine.
The Controversy
Some people are doubtful about how useful this new discovery will be. One of these individuals is Bruce Gelb, director of the Mindich Child Health and Development Institute. He says he simply can’t fathom that some people would use this technology instead of getting an accurate genetic test done. He’s steadfast in this belief even for lower income countries as genetic testing is becoming more widely available.
Dr. Gelb also says that the study conducted by this team may not have been completely accurate as it only utilized children with Noonan syndrome. The most prominent facial features associated with the condition diminish by adulthood, meaning their algorithm may only have the success rate researchers describe in adolescent patients.
The Positives
That all said, researchers in the field don’t deny that this is an impressive invention. They’re simply apprehensive about what tangible benefits might stem from it. But, others see some positive outcomes potentially developing from the algorithms.
First, the algorithms could be helpful for providing health care professionals who don’t have extensive training in rare genetic conditions more information more quickly.
Secondly, this development could help clinicians decide which genes need to be tested. This would better inform the lab tests they choose to order. Understanding which phenotype is being affected could ultimately help clinicians improve their ability to diagnose. That’s something that’s currently extremely difficult to do on ones own.
You can read more about this study and different professional opinions on it here.
