SDM

SDM predicts the effects of point mutations on protein stability using a knowledge-based statistical approach and computes a stability difference score from PDB coordinates and specified mutations.

Key Features:

• Knowledge-Based Approach: sdm2 employs an improved version of the Statistical Coupling Analysis (SDM) method, relying on statistical analysis rather than machine learning to assess mutation effects.
• Enhanced Substitution Tables: Environment-specific amino acid substitution tables are derived from an expanded Protein Data Bank (PDB) dataset reflecting a five-fold increase in available data.
• New Structural Parameters: The method incorporates packing density and residue depth as parameters for evaluating structural implications of mutations.
• Benchmark Testing: Validation was performed on a benchmark dataset comprising 2,690 point mutations from 132 protein structures.
• Stability Difference Score Calculation: Given a PDB file and a specified point mutation, the server calculates a stability difference score quantifying the mutation's impact on protein stability.

Scientific Applications:

• Drug Design: Predicting mutation-induced stability changes to inform design of compounds that stabilize or inhibit target proteins.
• Disease Modeling: Assessing the structural impact of genetic variants to support modeling of mutation-driven diseases.
• Protein Engineering: Identifying stabilizing or destabilizing substitutions to guide design of proteins with desired properties.

Methodology:

Utilization of an expanded PDB to refine environment-specific substitution tables; incorporation of packing density and residue depth parameters; application of an improved Statistical Coupling Analysis (SDM) knowledge-based statistical approach to predict mutation effects and compute stability difference scores, with validation against reverse mutations and a benchmark set of 2,690 point mutations from 132 structures.