MobiDB-lite

MobiDB-lite predicts long intrinsically disordered regions (IDRs) within protein sequences for proteome annotation and analysis of intrinsic disorder.

Key Features:

• Consensus-based prediction: Integrates outputs from eight predictors into an optimized consensus for IDR detection.
• Short-prediction filtering: Refines the consensus by filtering out spurious short predictions to prioritize extended regions.
• Improved specificity: Enhances specificity and reduces false positives for long disordered regions compared to individual methods.
• Targeting extended IDRs: Specifically focuses on detecting extended/long IDRs comparable in size to structured domains.
• Database integration: Integrated into MobiDB, DisProt, and InterPro to support large-scale proteome annotation.

Scientific Applications:

• Proteome annotation: Annotating proteomes with long IDR regions for genome- and proteome-scale analyses.
• Functional disorder studies: Investigating functional implications of intrinsic disorder in proteins.
• Domain-scale disorder detection: Detecting extended disordered regions comparable in size to structured domains to reduce single-residue prediction errors.

Methodology:

Integrates outputs from eight predictors into an optimized consensus which is refined by filtering out spurious short predictions.