PiPred - a deep-learning method for prediction of π-helices in protein sequences

Canonical pi-helices are short, relatively unstable secondary structure elements found in proteins. They comprise seven or more residues and are present in 15% of all known protein structures, often in functionally important regions such as ligand-and ion-binding sites. Given their similarity to alpha-helices, the prediction of pi-helices is a challenging task and none of the currently available secondary structure prediction methods tackle it. Here, we present PiPred, a neural network-based tool for predicting pi-helices in protein sequences. By performing a rigorous benchmark we show that PiPred can detect pi-helices with a per-residue precision of 48% and sensitivity of 46%. Interestingly, some of the alpha-helices mispredicted by PiPred as pi-helices exhibit a geometry characteristic of pi-helices. Also, despite being trained only with canonical pi-helices, PiPred can identify 6-residue-long alpha/pi-bulges. These observations suggest an even higher effective precision of the method and demonstrate that pi-helices, alpha/pi-bulges, and other helical deformations may impose similar constraints on sequences. PiPred is freely accessible at: https://toolkit.tuebingen.mpg.de/#/tools/quick2d. A standalone version is available for download at: https://github.com/labstructbioinf/PiPred, where we also provide the CB6133, CB513, CASP10, and CASP11 datasets, commonly used for training and validation of secondary structure prediction methods, with correctly annotated pi-helices.