Scoring Matrices: PAM and BLOSUM

 

Scoring Matrices: PAM and BLOSUM

Scoring matrices play a vital role in protein sequence alignment by assigning scores to matches, mismatches, and gaps. Two widely used matrices for this purpose are PAM and BLOSUM.

1. PAM (Point Accepted Mutation) Matrices

PAM matrices predict the likelihood of amino acid substitutions over evolutionary time. They are based on observed mutations in closely related proteins.

·         Key Concepts:

PAM1: Indicates a matrix where 1% of amino acids mutate, meaning sequences are approximately 99% identical.

Higher PAM Values (e.g., PAM250): Used for more divergent sequences that have undergone more mutations.

·         Applications: PAM matrices are typically used for global sequence alignment.

Example of PAM250 Matrix (simplified):

     A   R   N   D   C

A   2  -2   0   0  -2

R  -2   6  -1  -2  -4

N   0  -1   2   2  -4

D   0  -2   2   4  -5

C  -2  -4  -4  -5  12

2. BLOSUM (Blocks Substitution Matrix)

 BLOSUM matrices are based on conserved regions of protein sequences, focusing on observed substitutions within these regions.

·         Key Concepts:

BLOSUM62: The most commonly used matrix, constructed from sequences that share no more than 62% similarity.

Lower BLOSUM Numbers (e.g., BLOSUM45): Suitable for more divergent sequences.

·         Applications: BLOSUM matrices are ideal for local sequence alignment, particularly for tools like BLAST.

Example of BLOSUM62 Matrix (simplified):

     A   R   N   D   C

A   4  -1  -2  -2   0

R  -1   5   0  -2  -3

N  -2   0   6   1  -3

D  -2  -2   1   6  -3

C   0  -3  -3  -3   9

 

Key Differences Between PAM and BLOSUM

Feature	PAM Matrices	BLOSUM Matrices
Purpose	Models evolutionary change over time	Measures substitutions in conserved regions
Sequence Similarity	Higher numbers for divergent sequences (e.g., PAM250)	Lower numbers for divergent sequences (e.g., BLOSUM45)
Used For	Global alignment	Local alignment
Common Matrix	PAM250 for divergent sequences	BLOSUM62 for moderately divergent sequences

 

Applications of Scoring Matrices

1. Sequence Alignment: Scoring matrices help align nucleotide or protein sequences, rewarding matches and penalizing mismatches and gaps.
2. Phylogenetic Analysis: PAM and BLOSUM matrices can infer evolutionary relationships by comparing sequences.
3. Protein Function Prediction: Sequences with high alignment scores often have similar structures or functions, aiding in protein function prediction.
4. Database Searching: BLAST uses BLOSUM matrices, primarily BLOSUM62, to search databases for homologous genes or proteins.