Integrative Sequence-Based Classification Of Intrinsically Disordered Regions

Sequence-to-structure relationships have enabled the identification of conserved protein families and underly many of the classification approaches used to identify protein domains that fold as autonomous units. Intrinsically disordered proteins and regions (IDRs) represent domains that either fold upon binding or persist in disordered - albeit functional - states. IDRs are free from the constraints of autonomous folding and as a result their sequences are frequently less well conserved as assessed by conventional alignment-based metrics. Accordingly, traditional methods for sequence classification typically fail when trying to identify sequence families associated with IDRs. To address the challenge of demarking distinct subdomains within IDRs we have developed, ShanEnt, a Python toolkit for whole-proteome analysis of intrinsically disordered regions. ShanEnt uses sequence information to identify distinct types of IDRs by integrating over a collection of sequence features that include intrinsic features (those that can be extracted directly from the sequence) and extrinsic features (those identified using other methods). This approach allows for a robust and reproducible framework in which interactive exploration of proteomes can be performed via Jupyter notebook or customizable analysis pipelines. We have used ShanEnt to analyze a collection of proteomes, identifying distinct differences in IDRs across species and orthologous proteins. We hope ShanEnt will provide a useful approach to enable the effective classification and comparison of IDRs, with applications towards sequence design, evolutionary analysis, and ultimately the prediction of molecular functions in IDRs.