Effective High-Accuracy Prediction of Protein Structures from Easily Obtainable Artificial Homologous Sequences by Structure-Stability-Based Selection

Here, we developed an efficient, cost-effective assay for Structural Inference By Structure-stability-based selection of deep mutation variants and high throughput sequencing (Sibs-Seq). We demonstrated that unlike naturally occurring homologous sequences evolved over billions of years, these stability-selected, artificial homologs in a single-round 24 or 36 hours of mutation selections is often sufficient for producing high-accuracy structure prediction with < 2 angstrom root-mean-squared deviation (RMSD) to the native structure. And this is the processed data we generated for this work.