Complex models in cardiac electrophysiology are commonly used to simulate cardiac action potential and cardiac dynamics. However, their high complexity can lead to significant computational costs and challenges in large-scale simulations. Therefore, it could be interesting to find simplified versions of cardiac cell models that preserve the essential characteristics of the original system, enabling efficient simulations. Identification of simplified models must strike a balance between accuracy and complexity, ensuring that critical aspects in electrophysiology are preserved while maintaining simplicity and reduced computational costs. In this work, we explore the constrained formulation in PySINDy to identify cardiac electrophysiology models. The constrained feature allows the imposition of physically based conditions on the identified model terms and coefficients, ensuring that certain principles or mathematical properties are preserved. In particular, we demonstrate that polynomial formulations of the gating equations in the classical Hodgkin–Huxley model can be derived using the methodology developed in this work.