CoMemMoRFPred: Sequence-based Prediction of MemMoRFs by Combining Predictors of Intrinsic Disorder, MoRFs and...

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Molecular Recognition

Background:

• Intrinsically disordered regions (IDRs) contain molecular recognition features (MoRFs) that transition to ordered structures upon binding.
• Membrane-binding MoRFs (MemMoRFs) are functionally significant but lack dedicated computational prediction tools.
• Accurate identification of MemMoRFs is crucial for understanding their roles in cellular processes and disease.

Purpose of the Study:

• To develop and evaluate an accurate sequence-based computational predictor for membrane-binding MoRFs (MemMoRFs).
• To leverage existing state-of-the-art prediction tools for intrinsic disorder, lipid-binding, and disorder-to-order transitions.
• To provide a user-friendly webserver for predicting MemMoRFs to facilitate future research.

Main Methods:

• Utilized a dataset of experimentally annotated MemMoRFs for predictor development.
• Integrated predictions from three specialized tools: flDPnn (disorder), DisoLipPred (disordered lipid-binding), and MoRFCHiBi_Light (disorder-to-order transition).
• Applied a smoothing operator to refine predictions, mimicking native MemMoRF region characteristics.

Main Results:

• The combined approach, CoMemMoRFPred, demonstrated accurate prediction of MemMoRFs.
• The smoothing operator improved the accuracy and biological relevance of predicted MemMoRF regions.
• A publicly accessible webserver (http://biomine.cs.vcu.edu/servers/CoMemMoRFPred) was launched.

Conclusions:

• CoMemMoRFPred effectively predicts MemMoRFs by integrating multiple prediction strategies.
• The developed tool aids in exploring the abundance and functions of MemMoRFs.
• This predictor will support research into the roles of MemMoRFs in normal physiology and pathological conditions.