Mascot Server

Mascot Server performs protein identification and characterization from mass spectrometry data by integrating multiple search types and statistical scoring to match experimental spectra to protein sequences.

Key Features:

• Integrated Search Approach: Combines peptide molecular weights from enzymatic digestion, tandem mass spectrometry (MS/MS) data, and amino acid sequence data into a unified search framework.
• Probability-Based Scoring Algorithm: Uses a probability-based scoring algorithm that provides significance rules to minimize false positives, yields scores comparable with sequence homology searches, and supports iterative optimization of search parameters.
• Error-Tolerant Mode: Performs database matching of uninterpreted MS/MS data without enzyme specificity, accommodates chemical and post-translational modifications using a residue substitution matrix, and tests modifications serially to reveal additional peptide matches.
• Result Interpretation and Reporting: Employs a greedy set cover algorithm to generate a minimal set of proteins and hierarchical clustering with dendrograms to group proteins into families based on shared-peptide evidence.

Scientific Applications:

• High-throughput protein identification: Enables fully automated identification of proteins from mass spectrometry datasets using integrated mass, MS/MS, and sequence searches.
• Analysis of post-translational modifications: Detects and characterizes chemical and post-translational modifications in complex biological samples using error-tolerant searches and serial modification testing.
• Protein inference and characterization: Facilitates deriving minimal protein sets and assessing protein family evidence through probability-based scoring, greedy set cover selection, and hierarchical clustering with dendrograms.

Methodology:

Combines peptide mass (enzymatic digestion), MS/MS, and sequence searches; applies a probability-based scoring algorithm; performs error-tolerant database matching using a residue substitution matrix with serial modification testing; and uses a greedy set cover algorithm and hierarchical clustering to produce minimal protein sets and dendrograms.