Protein structure determines protein function. Despite the development of innovative laboratory techniques, experimental protein structure determination lags behind protein sequence accumulation. Computational methods seek to address this shortfall by predicting protein structure from sequence. Since 1994, protein structure prediction methods have been systematically assessed in the community-driven Critical Assessment of Protein Structure Prediction (CASP) competition. In recent CASP competitions, innovative machine learning (ML) models pushed the bounds of prediction to capture protein structures with experimental accuracy. ML has also brought rigor to protein design, increasing design success over traditional trial-and-error approaches. In this chapter, we review the basics of protein structure and the protein structure prediction problem. We examine prominent ML models from the most recent CASP13 to CASP15 competitions, followed by discussion of protein language and graph-based models for protein design.