5.4 Scoring and Evaluating Sequence Alignments

Scoring and evaluating sequence alignments is crucial for understanding the relationships between biological sequences. It involves assigning numerical values to matches, mismatches, and gaps, using substitution matrices like PAM and BLOSUM, and calculating overall alignment scores.

The choice of scoring scheme significantly impacts alignment outcomes and biological interpretations. Statistical measures like E-values and bit scores help assess alignment significance, while biological analysis of conserved regions and substitution patterns provides insights into evolutionary relationships and functional similarities.

Scoring Schemes for Alignment Quality

Numerical Scoring and Substitution Matrices

Top images from around the web for Numerical Scoring and Substitution Matrices

• 

  Chapter 3: Sequence Alignments – Applied Bioinformatics View original

  Is this image relevant?

• 

  Multiple sequence alignment - Wikipedia View original

  Is this image relevant?

• 

  SubVis: an interactive R package for exploring the effects of multiple substitution matrices on ... View original

  Is this image relevant?

• 

  Chapter 3: Sequence Alignments – Applied Bioinformatics View original

  Is this image relevant?

• 

  Multiple sequence alignment - Wikipedia View original

  Is this image relevant?

1 of 3

Top images from around the web for Numerical Scoring and Substitution Matrices

• 

  Chapter 3: Sequence Alignments – Applied Bioinformatics View original

  Is this image relevant?

• 

  Multiple sequence alignment - Wikipedia View original

  Is this image relevant?

• 

  SubVis: an interactive R package for exploring the effects of multiple substitution matrices on ... View original

  Is this image relevant?

• 

  Chapter 3: Sequence Alignments – Applied Bioinformatics View original

  Is this image relevant?

• 

  Multiple sequence alignment - Wikipedia View original

  Is this image relevant?

1 of 3

• Scoring schemes assign numerical values to matches, mismatches, and gaps in sequence alignments quantifying similarity between aligned sequences
• Substitution matrices (PAM and BLOSUM) provide pre-calculated scores for amino acid substitutions based on evolutionary relationships and biochemical properties
• Overall alignment score calculated by summing individual position scores measures alignment quality and sequence similarity
• Gap penalties incorporated to account for insertions and deletions with distinctions between gap opening and gap extension penalties
  • Gap opening penalty typically higher than gap extension penalty
  • Example: Gap opening penalty of -10, gap extension penalty of -2
• Different scoring schemes optimized for various sequence types and evolutionary distances
  • DNA scoring schemes often use simple match/mismatch scores (1 for match, -1 for mismatch)
  • Protein scoring schemes utilize more complex substitution matrices

Impact of Scoring Choices

• Choice of scoring scheme and substitution matrix significantly impacts alignment outcome and biological interpretation
  • Conservative scoring favors exact matches leading to shorter alignments
  • Permissive scoring allows more mismatches resulting in longer alignments
• Scoring schemes tailored to specific sequence types
  • DNA/RNA: Simple match/mismatch schemes
  • Proteins: Complex substitution matrices accounting for amino acid properties
• Evolutionary distance consideration crucial for scoring scheme selection
  • Closely related sequences: Higher penalties for mismatches and gaps
  • Distantly related sequences: More permissive scoring allowing for more substitutions

Choosing Scoring Matrices

PAM and BLOSUM Matrices

• PAM (Point Accepted Mutation) matrices suitable for closely related sequences
  • Lower PAM numbers indicate shorter evolutionary distances
  • Example: PAM1 for very closely related sequences, PAM250 for more distant relationships
• BLOSUM (Blocks Substitution Matrix) matrices appropriate for more distantly related sequences
  • Higher BLOSUM numbers suitable for closer relationships
  • Example: BLOSUM62 widely used for general protein sequence alignments
• Choice between PAM and BLOSUM matrices depends on estimated evolutionary distance between sequences
  • PAM matrices based on global alignments of closely related proteins
  • BLOSUM matrices derived from local alignments of more divergent sequences

Specialized Matrices and Gap Penalties

• Protein-specific matrices incorporate amino acid properties
  • BLOSUM62 accounts for biochemical similarities between amino acids
  • Example: Higher score for leucine-isoleucine substitution compared to leucine-lysine
• Custom scoring matrices necessary for specialized applications or non-standard sequence types
  • Transmembrane proteins require matrices accounting for hydrophobicity
  • Repetitive DNA elements need scoring schemes handling repeat structures
• Gap penalties adjusted based on expected frequency and length of insertions/deletions
  • Lower gap penalties for more divergent sequences allowing more indels
  • Higher gap penalties for closely related sequences favoring fewer gaps
• Selection of matrices and gap penalties often requires empirical testing and optimization
  • Iterative refinement based on alignment quality and biological relevance
  • Benchmarking against known homologous sequences or structures

Statistical Significance of Alignments

E-values and Bit Scores

• E-values (expectation values) represent number of alignments with given score expected to occur by chance in database of particular size
  • Lower e-values indicate higher statistical significance
  • Values typically below 1e-5 or 1e-10 considered significant for most applications
• E-value calculation depends on factors such as sequence length, database size, and specific scoring system used
  • Longer sequences or larger databases increase chance of random matches
  • More stringent scoring systems result in lower e-values for same raw score
• Bit scores provide normalized measure of alignment quality independent of database size
  • Allows comparison of alignments across different searches
  • Calculated using Karlin-Altschul statistics

Additional Significance Measures

• Z-scores measure number of standard deviations alignment score deviates from mean of random distribution of scores
  • Higher Z-scores indicate greater statistical significance
  • Example: Z-score of 3 means alignment score is 3 standard deviations above mean
• Karlin-Altschul statistics provide theoretical framework for assessing significance of local sequence alignments
  • Based on extreme value distribution of local alignment scores
  • Used to calculate e-values and bit scores
• Multiple hypothesis testing corrections necessary when evaluating significance of multiple sequence alignments simultaneously
  • Bonferroni correction adjusts p-value threshold based on number of comparisons
  • Example: For 100 comparisons, significance threshold lowered from 0.05 to 0.0005

Biological Implications of Alignments

Evolutionary and Functional Insights

• High-scoring alignments may indicate evolutionary relationships, functional similarities, or conserved structural elements between sequences
  • Example: High similarity in catalytic domains of enzymes suggests conserved function
• Distribution and pattern of conserved regions within alignment provide insights into functionally important domains or motifs
  • Example: Highly conserved zinc finger motifs in DNA-binding proteins
• Gaps in alignments may represent insertions, deletions, or regions of low sequence conservation
  • Potentially indicating functional divergence or structural flexibility
  • Example: Large insertion in one sequence may represent a novel functional domain

Critical Analysis and Interpretation

• Presence of specific amino acid substitutions can suggest functional adaptations or constraints on protein evolution
  • Conservative substitutions (leucine to isoleucine) often maintain function
  • Radical substitutions (glycine to tryptophan) may indicate functional changes
• Alignment quality and significance measures considered in conjunction with biological context and experimental evidence
  • High-scoring alignment not always indicative of biological relevance
  • Low-scoring alignment may still be biologically significant in certain contexts
• Multiple sequence alignments reveal patterns of conservation across species or protein families
  • Informing evolutionary and functional analyses
  • Example: Identifying catalytic residues conserved across enzyme families
• Critical analysis considers potential limitations of alignment method
  • Handling of repetitive elements
  • Impact of compositional bias in sequences
  • Alignment artifacts in regions of low complexity