Bioinformatics: Protein Sequence Classification

Classifying Sequences

• Sequences are classified to predict their function.

Levels of Classification

• Classification occurs at various levels: families, domains, motifs, patterns, profiles, and fingerprints.

Classification Tools

• InterProScan is used as a classification tool.

BLAST

• BLAST can identify unknown sequences with homologous sequences in databases revealing information about its biology.
• Sequences with similar structures and features are likely to have similar structure and function.

Protein Classification

• Proteins are classified based on families, domains, and sequence features.
• Grouping new proteins with other sequences is possible due to common features.

Protein Families

• A group of proteins share a common evolutionary origin.
• This is reflected in both their related functions and sequence or structural similarities.
• As families get smaller their members have more in common.

Protein Domains

• Domains are functional and/or structural units in a protein, responsible for a particular function or interaction, contributing to the overall role of a protein.
• Similar domains are found in proteins with different functions.
• Cystatin domains are typically 115 amino acids long and function as protease inhibitors
• Cystatin domains feature conserved inhibitory motifs, a wedge-like shape, and block the active site of proteases.
• Proteins can have multiple domains, each with specific roles.
• Family and domain-based classifications can overlap.
• Regulator of G-protein Signalling (RGS) domains are found in proteins from different families.

Sequence Features

• Sequence features are small groups of amino acids.
• They confer some biochemical property upon a protein and usually are linked to overall function.
• Sequence features can occur within protein domains.

Sequence Features Examples

• Active sites are catalytic residues of enzymes; serine proteases have an active site triad of His, Asp, Ser.
• Binding sites are residues that directly bind molecules or ions such as tubulin having GTP binding sites.
• Post-translational modification (PTM) sites are chemically modified residues, examples including glycosylation, phosphorylation, and oxidation.
• Glycosylation, Phosphorylation, Methylation and Acetylation represent a Post-translational modification (PTM).

Types of Sequence Features

• Sequence features can be classified as Motifs, Patterns, Profiles or Signals.

Sequence Motifs

• Motifs are short, conserved regions of amino acid or DNA sequences.
• Motifs are structurally and functionally important.
• Motifs contain biologically important sequences.
• An "LSH" motif is required for activation of cysteine proteases.

Sequence Patterns

• Describes a motif using a qualitative consensus sequence, identified from multiple sequence alignments, for example N-{P}-[ST]-{P}
• N-{P}-[ST]-{P} is interpreted as N-{any except P}-[S or T]-{any except P}

Sequence Profiles

• Describes motifs using quantitative data captured in position-specific scoring matrix.
• Quantitative data uses substitution matrices like those used to score sequence alignments.

Signal Sequences

• Signal sequences deliver proteins to certain sites within or outside the cell.
• Proteins are produced in the cytoplasm.
• Proteins move to a specific place depending on the signal sequence.
• The TargetP tool predicts the cellular location of proteins. It predicts the cellular location of proteins based on the presence of signal sequences.
• Signal peptides are short sequences that direct proteins for secretion outside the cell.
• SignalP server predicts the presence and location of signal peptide cleavage sites.

Sequence Fingerprints

• Many proteins contain more than one sequence motif.
• Fingerprints refer to a collection of motifs that occur in a specific order.

InterPro

• It brings patterns, profiles, and fingerprints from different databases together into a single searchable resource.
• It is useful for identifying conserved regions and motifs in proteins and viewing them.