Sites Inferred by Metabolic Background Assertion Labeling (SIMBAL): adapting the Partial Phylogenetic Profiling...

Area of Science:

• Genomics
• Bioinformatics
• Structural Biology

Background:

• Comparative genomics and phylogenetic profiling are powerful tools for analyzing biological data.
• Existing methods can be limited by data noise and the need for extensive training sets.

Purpose of the Study:

• Introduce SIMBAL (Sites Inferred by Metabolic Background Assertion Labeling), a novel method for identifying short functional sequence signatures.
• Validate SIMBAL's ability to find functionally relevant signatures in protein families.

Main Methods:

• SIMBAL applies Partial Phylogenetic Profiling (PPP) locally within protein sequences.
• It uses binomial distribution statistics to optimize similarity cutoffs on partitioned training sets.
• The method involves sliding windows across protein sequences and searching subsequences against partitioned datasets.

Main Results:

• SIMBAL successfully identified short sequence signatures correlated with specific functional traits in ABC permeases (urea utilization) and protein methyltransferases (substrate RF-1 proximity).
• For ABC permeases, identified sites were gating determinants on the cytosolic face.
• For methyltransferases, identified sites mapped to the substrate-binding interface, distinct from the reaction mechanism regions.

Conclusions:

• Short sequence signatures identified by SIMBAL can outperform full-length sequences for functional predictions.
• SIMBAL provides a method for localizing and modeling functional sites, especially when integrated with structural data.