Local and Global Alignment

 

Local and Global Alignment

 

Local Alignment

Local alignment focuses on finding regions of high similarity within two sequences. It aligns only the portions of the sequences that are most similar while disregarding the rest. This method is particularly useful when the sequences have significant differences in length or when only certain regions are expected to be homologous.

 

Ø  The Smith-Waterman algorithm, used for local alignment, identifies the highest scoring subsections of the sequences based on a scoring matrix.

Ø  It starts by assigning scores to individual matches or mismatches between residues and penalizing gaps.

Ø  Instead of forcing alignment of the entire sequence, the algorithm halts alignment at positions where the score becomes negative, ensuring that only the most relevant regions are aligned.

Example:

Consider two sequences:

• Sequence 1: ACGTACGT
• Sequence 2: TACGTGAC

Local alignment might identify the subsequence ACGT as the most similar region between the two sequences.

 

Global Alignment

Global alignment involves aligning two sequences from start to end, including all residues in both sequences. It seeks to maximize the overall similarity score across their entire length. This approach is ideal when comparing sequences of similar lengths and compositions.

How it Works:

Ø  The Needleman-Wunsch algorithm, used for global alignment, constructs a scoring matrix to evaluate all possible alignments between the sequences.

Ø  It ensures that gaps or mismatches are accounted for by applying penalties but includes them in the alignment to produce an optimal match.

Ø  The entire length of both sequences is aligned, even if it results in regions with lower similarity scores.

Example:

Consider the same sequences:

• Sequence 1: ACGTACGT
• Sequence 2: TACGTGAC

Global alignment might produce:

 

-ACGTACGT

 TACGTGAC-

This aligns the sequences completely, introducing gaps as needed.

 

Comparison of Local and Global Alignment

Aspect	Local Alignment	Global Alignment
Objective	Find regions of similarity	Align entire sequences
Algorithm	Smith-Waterman	Needleman-Wunsch
Output	Partial alignment	Full alignment
Best for	Dissimilar or variable-length sequences	Similar or closely related sequences
Applications	Domain/motif search	Comparative genomics
Flexibility	High	Moderate