AI aids efforts in ALS

PHOENIX—Researchers at Barrow Neurological Institute have validated an earlier study, using the artificial intelligence (AI) of IBM’s Watson to verify the discovery of five new genes linked to amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s disease.

The latest study results, validated through five different methods, were published in November 2017 in Acta Neuropathologica, in an online manuscript entitled, “Artificial intelligence in neurodegenerative disease research: Use of IBM Watson to identify additional RNA-binding proteins altered in amyotrophic lateral sclerosis.”

“The significance of our work is that it demonstrates that the use of artificial intelligence methodologies for scientific discovery can greatly accelerate the more traditional laboratory based research approach,” says Robert Bowser, a professor of neuroscience at Barrow Neurological Institute.

Watson uses machine learning, natural language processing and other cognitive reasoning technologies and was trained about proteins that researchers already knew were related to the disease using data, as well as all published research information surrounding ALS.

“The identification of additional genes/proteins for ALS can take years since we don’t know where to focus our research efforts,” Bowser says. “And the use of Watson for Drug Discovery quickly provided insight into which genes/proteins to further explore and validate.”

Considered one of the nation’s top ALS researchers, Bowser, along with his team, provided a list of 11 RNA binding proteins with known mutations that cause ALS. Watson used the list of proteins and cross-referenced medical literature from 28 million MEDLINE abstracts to rank order all other 1,500 RNA binding proteins encoded by the genome to attempt to identify new RNA binding proteins linked to ALS.

The research team validated the top 10 RNA binding proteins using five different methods that included use of patient tissue samples and patient-derived stem cells differentiated into motor neurons. Researchers also examined a smaller set of RNA binding proteins near the bottom of the list to demonstrate that any changes detected in the top 10 were not observed for those at the bottom of the list, demonstrating the ability of Watson to correctly predict RNA binding proteins linked to ALS.

The results were “groundbreaking,” according to Bowser.

“Further validating and expanding on our earlier findings has been exciting because in research of this nature, time is of the essence,” Bowser tells DDNews. “We could have individually looked at the 1,500 proteins and genes, but it would have taken us much longer to do so. These findings inspire hope that, with this technology, we may someday identify new and more effective treatments for ALS.”

In 2011, IBM Watson showed off its skills by competing against human contestants on Jeopardy—and won. Watson for Drug Discovery launched in August 2016, working with clients and partners on a variety of research centered on neurodegenerative diseases such as ALS, Parkinson’s disease, Alzheimer’s disease and more. On Dec. 14, 2016, Barrow announced that Watson had performed a “revolutionary study” which identified new genes linked to ALS discovery, giving researchers new insights. This study was validated the following November.

A fatal neurological disorder affecting more than 220,000 patients worldwide, ALS has no cure and few treatments. It is a progressive and degenerative disease in which the cells that control voluntary muscle movements die, leading to paralysis and ultimately, death, Bowser explains. Scientists don’t yet know what causes ALS, and there is just one FDA-approved medication for the disease that is only marginally effective. Most individuals die from ALS within three to five years from the onset of symptoms, and approximately 6,000 people are diagnosed with ALS every year.

More than 30 genes have been linked to ALS, and mutations in the 11 genes that encode RNA binding proteins cause familial forms of ALS. These RNA binding proteins play a critical role in how genes encoded within the DNA in every cell are converted to the proteins that perform all the functions within a cell.

“We are currently performing additional laboratory research studies to explore how the top genes/proteins identified by Watson contribute to motor neuron degeneration,” Bowser comments. “In addition, we are exploring whether any genetic alterations of the Watson identified RNA binding proteins occur in the ALS patient population.

“The goal is to use our new knowledge to define specific molecular mechanisms that can be targeted for drug development.”

The Barrow laboratory studied whether IBM Watson could accelerate the identification of additional RNA binding proteins (RBPs) linked to ALS by helping scientists focus research efforts on the proteins that Watson ranked high and predicted to be altered in ALS.

“Overall, we successfully used IBM Watson to help identify additional RBPs altered in ALS, highlighting the use of artificial intelligence tools to accelerate scientific discovery in ALS and possibly other complex neurological disorders,” the authors of the Acta Neuropathologica manuscript conclude.