Introduction: ermodynamic Stability, Structure, and Disorder 206 Historical View of Intrinsic Disorder as an All-or-None Phenomenon 206 An Alternative View of Disorder as Part of an Energetic Continuum 207

Modeling ermodynamic Stability of Structured and Disordered Regions in Proteins 208

Prediction of ermodynamics from Sequence Using the eScape Algorithm 208 Analyzing the Energetics of Structured and Disordered Residues 211 Modeling the Energetics of Residue Stability with Normal Distributions 211 Intrinsically Disordered Residues Exhibit Lower Predicted Stability 214

Basic Native State Stability Predictor of Disorder from Amino Acid Sequence 215 Di erences in Mean Stability Correlate with Disorder Propensity and Physical Property Scales 215 Development of a Stability-Based Probability Model for Disorder Prediction 217

More Sophisticated Support Vector Machine Predictors of Disorder Based on Multiple ermodynamic Parameters 222

Support Vector Machine Predictor Based on Native State ΔG Alone 222 SVM Predictor Based on Eight Native and Denatured State ermodynamic Quantities 223

Conclusions and Outlook: e Importance of Protein ermodynamics for Intrinsic Disorder 227 Acknowledgments 228 References 228

e rst crystal structures of protein molecules represented breakthroughs of immense importance that transformed our collective understanding of the physical processes sustaining life (Kendrew et al. 1958; Muirhead and Perutz 1963). An unintended side e ect of those discoveries, however, was the cementing of the biological paradigm of “one gene, one protein, one structure, one function” (Dickerson 1972; Lehninger 1975). Further investigation over the course of decades demonstrated that this notion was insucient to describe all aspects of the molecular function of proteins. Examples where this reigning paradigm presently fails include (but are not limited to) dynamic allostery (Cooper and Dryden 1984; Tzeng and Kalodimos 2011), protein as ensemble (Boehr et al. 2009; Hilser et al. 2006), functional moonlighting (Copley 2012; Je ery 2009), metamorphic proteins (Bryan and Orban 2010; Murzin 2008), frameshift alternative translation (Michel et al. 2012), and intrinsically disordered proteins (Dunker and Obradovic 2001; Uversky 2013).