Misprediction of Structural Disorder in Halophiles

Area of Science:

• Biochemistry
• Bioinformatics
• Structural Biology

Background:

• Intrinsic disorder in proteins is typically studied using biophysical methods, but proteome-scale analyses rely on bioinformatics predictions.
• The accuracy of these prediction tools for proteins from extremophilic microorganisms, adapted to harsh environments, is poorly understood.
• Extremophiles may exhibit sequence composition biases due to adaptation, potentially affecting disorder prediction performance.

Purpose of the Study:

• To evaluate the performance of common protein disorder prediction tools on extremophile proteomes.
• To compare disorder predictions across different extremophile groups (hyperthermophiles, radiotolerants, halophiles) and a reference mesophilic group.
• To identify potential biases and mispredictions in disorder prediction for extremophilic proteins.

Main Methods:

• Predicted structural disorder for complete proteomes of various extremophile groups using established bioinformatics tools.
• Compared prediction results against a reference mesophilic proteome.
• Analyzed Protein Data Bank (PDB) entries to validate predictions and understand biophysical principles of extremophilic protein behavior.

Main Results:

• Significant deviations in disorder predictions were observed between extremophiles and mesophiles.
• Hyperthermophiles showed a lack of disordered regions, while radiotolerants exhibited a gain.
• Halophiles demonstrated systematic overprediction of disorder, with PDB examples revealing specific surface properties contributing to misprediction.

Conclusions:

• Current bioinformatics tools show limitations in accurately predicting protein disorder in extremophiles, particularly halophiles.
• Understanding the unique biophysical properties of extremophilic proteins is crucial for interpreting disorder predictions.
• While dedicated predictors are currently infeasible due to data scarcity, recommendations are provided for optimizing disorder analysis with existing tools.