Researchers at Emory University have identified a panel of 10 proteins that distinguished patients with amyotrophic lateral sclerosis (ALS) from healthy individuals with 98.3% accuracy, according to a study published in Nature. The findings could contribute to efforts to identify the progressive neurological disease, which causes the gradual loss of voluntary muscle control and eventually leads to paralysis.
ALS, also known as Lou Gehrig’s disease, is a fatal neurodegenerative condition that affects motor neurons in the brain and spinal cord. These nerve cells control voluntary movements, and their deterioration progressively limits patients’ ability to move. Diagnosis typically occurs between the ages of 40 and 70. Nearly 90% of cases are sporadic, meaning patients have neither a family history of the disease nor an identified genetic mutation associated with it.
The absence of an identifiable cause in most cases has encouraged researchers to investigate biological markers that could help distinguish ALS through non-invasive testing, including blood analysis.
Eleanor Thomas, M.D., Ph.D., an assistant professor of neurology at Emory University and the study’s corresponding author, discussed the potential clinical implications of the findings in an interview with Fierce. She said, “The clinical impact could be getting patients to ALS centers sooner,” highlighting the possibility of reducing delays before patients reach these centers.
Researchers narrow down the protein panel
To identify proteins associated with ALS, the research team collected blood samples from 632 patients diagnosed with the disease and 160 healthy individuals. The samples were examined using protein panels containing 319 markers associated with the central nervous system and inflammation.
The researchers subsequently applied machine learning to determine which proteins provided the strongest signals for differentiating ALS patients from healthy controls. This process reduced the initial selection to 12 proteins, which were then tested independently using another group comprising 144 ALS patients and 37 healthy individuals.
Following this validation process, the researchers established a final panel containing 10 proteins. Its ability to distinguish patients with ALS from healthy controls reached 98.3% accuracy.
The team also examined how the number of proteins affected the panel’s performance. Although accuracy increased as additional proteins were incorporated, the improvement levelled off once the selection reached 10. Including further proteins did not produce additional gains in accuracy.
Biomarkers reflect different aspects of ALS
Some proteins included in the final panel had already been investigated as ALS biomarkers. These included pTau-181, a phosphorylated form of tau protein, as well as neurofilament light chain (NEFL) and neurofilament heavy chain (NEFH).
Thomas described neurofilament as the “best ALS biomarker,” but noted that it does not identify every patient. The remaining proteins in the panel did not overlap with previously detailed biomarker panels for ALS.
These additional proteins are associated with several aspects of ALS biology, including inflammation, neuronal damage, vascular dysfunction and cellular stress.
Further investigation planned
Thomas and her colleagues intend to conduct additional research using patient samples that more closely represent the individuals encountered in clinical practice. This will include patients whose condition is initially misidentified as myasthenia gravis.
Thomas also suggested that some biomarkers useful for distinguishing diseases could eventually provide information about disease prognosis. The researchers plan to investigate this possibility as they examine the identified proteins in more representative patient samples.
Emory Researchers Identify a 10-Protein Panel for ALS Detection
Researchers at Emory University School of Medicine have identified a combination of 10 blood proteins that can distinguish people with amyotrophic lateral sclerosis (ALS) from healthy individuals with 98.3% balanced accuracy. The discovery could support the development of a blood-based screening tool to help recognize ALS earlier.
The findings were published in Nature Communications on October 19, 2026. The research represents an important step toward improving the identification of a progressive neurological disease that can be difficult to diagnose during its early stages.
Emory Researchers Plan Further Validation
Although the initial findings are promising, the research team recognizes that more work is required before the panel can become a routine clinical tool.
Future studies will need to include patients with other neurological and neuromuscular conditions that can resemble ALS. Researchers must determine whether the protein combination can reliably distinguish these conditions in real-world diagnostic settings.
The team also aims to investigate additional patient samples that reflect the complexity of clinical practice, including cases in which ALS may initially be mistaken for another disorder.
Conclusion: Emory Advances the Search for Better ALS Detection
The Emory study highlights the potential of combining multiple blood-based protein biomarkers with machine learning to improve ALS identification. Its 10-protein panel achieved 98.3% balanced accuracy when distinguishing people with ALS from healthy individuals.
However, the protein panel is not yet an approved standalone diagnostic test. It distinguished ALS patients from healthy controls in the study, and additional research is needed to establish how well it performs against other neurological conditions with similar symptoms.
Emory Research Could Support Future ALS Clinical Trials
Clinical trials depend on identifying appropriate participants and accurately understanding their disease characteristics. Delays in diagnosis can make recruitment and eligibility assessment more difficult.
The Emory protein panel could eventually support clinical research by providing another source of biological information during patient evaluation. Researchers may also investigate whether these proteins can help predict disease progression or identify different biological subgroups of ALS.
If future studies confirm its usefulness, the approach could help clinical research teams improve participant selection and investigate new treatment strategies.


