NetraMark Says Its Generative AI Discovers Novel Rare Disease Drug Targets

Rare Daily Staff

Generative AI developer NetraMark said newly published research shows its NetraAI detected novel and known potential therapeutic targets for the rare, neurodegenerative disease amyotrophic lateral sclerosis that were not identified using other AI-based methods.

The study, published in the journal Frontiers in Computational Neuroscience, also said NetraAI identified well-defined subpopulations that could improve ALS clinical trial outcomes. The results were derived from 116 patients, which the company said demonstrates the power of its platform compared with other AI methods that require large data inputs.

The data identified several genes that shed light into ALS pathophysiology and represent new avenues for treatment. The analysis also identified subpopulations of ALS patients based on disease onset.

In the study, researchers used NetraAI to analyze data collected by Answer ALS, the largest collaborative effort in ALS, bringing together multiple research organizations and key opinion leaders. Over 800 ALS patients and 100 healthy controls from eight neuromuscular clinics distributed across the United States were enrolled in this project. NetraAI was available to medical experts at the Gladstone Institute, allowing them to interact with the ML-generated hypotheses and to evaluate the findings and examine the causal factors that the NetraAI model suggested. This approach bridges a critical gap that exists between advanced ML techniques and human medical expertise.

In contrast with other AI-based methods for analyzing the data set, NetraAI is engineered to include focus mechanisms that separate small datasets into explainable and unexplainable subsets. Unexplainable subsets are collections of patients that can lead to suboptimal overfit models and inaccurate insights due to poor correlations with the variables involved.

The NetraAI uses the explainable subsets to derive insights and hypotheses (including factors that influence treatment and placebo responses, as well as adverse events) that can significantly increase the chances of a clinical trial success. State of the art AI methods lack these focus mechanisms and assign every patient to a class, even when this leads to so-called “overfitting,” which drowns out critical information that could have been used to improve a trial’s chance of success.

“These data and related insights underscore the differentiated capabilities of NetraAI compared with other AI-based solutions for target discovery and clinical trial analytics,” said Joseph Geraci, chief technology officer, chief scientific officer, and co-founder of NetraMark, and first author on the publication. “Using a small ALS dataset and a unique machine learning paradigm, we have not only validated previously reported ALS drug targets but also uncovered critical insights into ALS heterogeneity.”

Geraci said a major problem with state-of-the-art AI methods for understanding patient populations is that they lack the ability to be critical of how the data is labeled. He said people are very complex systems, and the NetraAI system is specifically geared to uncover aspects of patient populations that go beyond human observation and expose various driving factors behind the biology of the disease.

“This technology provides our company with unique insights into how people are truly being affected by a disease and has the potential to enable our clients to significantly improve clinical trial outcomes,” he said.