ISDTool

ISDTool predicts immunosuppressive domains (ISDs) in retroviral envelope transmembrane (env) proteins to identify regions that modulate immune responses and facilitate viral immune escape.

Key Features:

• Biological target: Detects ISDs within transmembrane proteins of the envelope gene (env), regions implicated in inhibiting lymphocyte proliferation and enabling immune escape.
• Retrovirus scope: Applies to human endogenous retrovirus (HERV), HTLV-1, HIV-1, STLV, SIV and six typical retroviruses from three species.
• Computational model: Integrates minimum Redundancy Maximum Relevance (mRMR) feature selection with a weighted extreme learning machine (WELM) classifier.
• Validation performance: Reports identification accuracies of 98.95% (self-consistency), 96.34% (5-fold cross-validation), and 96.87% (10-fold cross-validation).
• Motif and evolutionary analysis: Predicts putative ISDs and analyzes motifs across retroviruses to provide insights into evolutionary relationships.

Scientific Applications:

• Immune evasion studies: Identification of ISDs to investigate mechanisms by which retroviruses inhibit lymphocyte proliferation and evade innate and adaptive immune effectors.
• Comparative evolutionary analysis: Motif-based comparison of ISDs across HERV, HTLV-1, HIV-1, STLV, and SIV to study evolutionary relationships.
• Pathogenesis research: Characterization of ISDs to inform studies on retroviral immunosuppression and its implications in viral pathogenesis.

Methodology:

The method combines minimum Redundancy Maximum Relevance (mRMR) feature selection with a weighted extreme learning machine (WELM), includes motif analysis of predicted ISDs, and evaluates performance using self-consistency, 5-fold, and 10-fold cross-validation.