Identifying scale-invariant geometric features in high-dimensional data remains a central challenge in data science. While the Mapper algorithm offers a powerful topological lens, its utility is often hindered by a critical dependence on manual parameter tuning and a lack of reproducibility. Conversely, persistent homology provides robust multiscale insights but lacks a direct, automated integration into structural discovery.

​To bridge these gaps, we propose a fully automated topological framework that unifies Mapper with persistent homology to achieve principled, noise-resilient feature extraction. The core innovation is our Most Persistent Betti-1 (MPB-1) algorithm, which systematically extracts dominant one-dimensional homological features from persistence diagrams to compute a characteristic scale (ε). This scale guides the parameterization of the Mapper algorithm through a closed-form relationship, eliminating traditional trial-and-error.

​We validate our framework on synthetic datasets, where it accurately recovers theoretical Betti-1 invariants, and on real-world applications including single-cell genomics and protein dynamics. Our results reveal robust topological signatures consistent with underlying biological structures. To our knowledge, this is the first automated Mapper framework that guarantees fidelity to persistent topological invariants. By embedding Betti-1 computation at its core, our method provides a reproducible foundation for detecting cyclic structures central to medicine and proteomics, thereby broadening the reach of topological data analysis across scientific domains.