An essential step in searching for a way to prevent neural tube congenital disabilities like spina bifida is understanding the defects’ genetic underpinnings. Scientists have long suspected genes play a significant role in developing spina bifida, but research has not yet been able to zero in on definite genetic causes. Now, researchers at Weill Cornell Medicine in New York have conducted a study that may help close some of the knowledge gaps that have hindered progress.
Spina bifida is a congenital disability in which an embryo’s central nervous system does not develop properly early in pregnancy. As a result, a baby with spina bifida is born with a spinal column that is not fully closed, leaving vital nerve tissue exposed and creating the risk of severe complications.
Spina bifida sometimes runs in families and is more common in some populations than in others. These factors lead scientists to suspect that genes play at least some role in the risk of developing spina bifida and other similar congenital disabilities. However, the connection between the condition and genes is not simple. Rather than being the result of a single gene mutation, the defect results from several different gene variations. External conditions during pregnancy also factor in.
Using Artificial Intelligence to Compare Genes
To find the genes involved, researchers have studied mice. They’ve made some progress, but their research has had some limitations. For example, as they’ve discovered genes in the mice implicating neural tube defects, they’ve gained clues to which human genes might be involved. But by focusing on those genes, they run the risk of overlooking other genes that might be critical pieces of the genetic puzzle.
The researchers at Weill Cornell Medicine had an idea for getting past that kind of research bias. Suppose they could somehow compare the genetic characteristics of many people with spina bifida, looking at their entire genetic makeup instead of only a few genes. In that case, they might be able to spot critical similarities without focusing solely on pre-determined genes.
The scientists recruited about 300 participants, both people with spina bifida and healthy controls, and subjected them to a genome-wide association study (GWAS). This kind of study compares the complete genomes of the participants. The goal is to find genetic commonalities between the people with the disorder distinguishing them from the control group.
The challenge of GWAS is that it requires processing and comparing vast amounts of genetic data. To overcome this hurdle, researchers employed artificial intelligence software tools, which used machine learning to identify genetic variations that might be of interest.
The results were promising. Among the genes flagged by the software were some involved in the body’s ability to use sugars and fats. These findings align with what researchers already know about neural tube defects.
“These processes are relevant to conditions like diabetes and obesity,” Dr. Margaret Elizabeth Ross, the study’s senior author, told the Cornell Chronicle. “This gave us a lot of encouragement that our machine learning approach was coming up with clinically relevant information.”
While the results are preliminary, they could be the first step in understanding the genetic mechanism of spina bifida and identifying ways to prevent the condition from developing.