FISH - profile search

FISH - profile search identifies the family membership of domains within query protein sequences using Structure Anchored Hidden Markov Models (saHMMs) to enable accurate domain classification even at low sequence identity.

Key Features:

• High accuracy and performance: Evaluation on SCOP sequences with an E-value cut-off of 0.1 reported that 99.3% of top hits correspond to the correct saHMM.
• Functional annotation and structural insights: Provides domain annotations, predicted probable two-dimensional (2D) and three-dimensional (3D) domain structures, and potential functional information for matched domains.
• Sequence alignment capabilities: Performs pairwise and multiple sequence alignments to homologues exhibiting low sequence identity to support comparative and evolutionary analyses.

Scientific Applications:

• Protein domain classification: Assigns domains to families to aid inference of protein function and domain membership.
• Structural biology research: Supplies predicted 2D and 3D domain structures to support studies of protein architecture and dynamics.
• Evolutionary studies: Enables exploration of evolutionary relationships among proteins with low sequence identity through alignment to distant homologues.

Methodology:

Uses Structure Anchored Hidden Markov Models (saHMMs) that integrate structural information for domain identification and performs pairwise and multiple sequence alignments to homologues with low sequence identity.