The study introduces MuFaDDG, a sequence-based tool for predicting protein stability changes. Frataxin was used as the primary case study (via the CAGI5 challenge) to validate the model's performance. MuFaDDG achieved high accuracy (ACC: 0.81) in predicting how mutations affect frataxin's stability, outperforming existing state-of-the-art methods.Conclusion: The model proves highly effective at identifying destabilizing mutations in frataxin, which is critical for understanding diseases like Friedreich's Ataxia.