ElixirSeeker: A Machine Learning Framework Utilizing Fusion Molecular Fingerprints for the Discovery of Lifespan-Extending Compounds

Abstract

Despite the growing interest in developing anti-aging drugs, high costs and low success rates of traditional drug discovery methods pose significant challenges. Aging is a complex biological process associated with numerous diseases, making the identification of compounds that can modulate aging mechanisms critically important. Accelerating the discovery of potential anti-aging compounds is essential to overcome these barriers and enhance lifespan and healthspan. Here, we present ElixirSeeker, a machine learning framework designed to maximize feature capture of lifespan-extending compounds through multi-fingerprint fusion mechanisms. Utilizing this approach, we identified several promising candidate drugs from external compound databases. We tested the top six hits in Caenorhabditis elegans and found that four of these compounds—including Praeruptorin C, Polyphyllin VI, Thymoquinone, and Medrysone—extended the organism's lifespan. This study demonstrates that ElixirSeeker effectively accelerates the identification of viable anti-aging compounds, potentially reducing costs and increasing the success rate of drug development in this field.