Related Experiment Videos
Aligning amino acid sequences: comparison of commonly used methods
Journal of Molecular Evolution
|January 1, 1984
Summary
Comparing protein sequence alignment methods, the Dayhoff log-odds matrix (LOM) generally best identifies distant evolutionary relationships, though simpler methods can suffice for closely related protein families.
Area of Science:
- Bioinformatics
- Computational Biology
- Molecular Evolution
Background:
- Protein sequence alignment is crucial for inferring evolutionary relationships.
- Various weighting schemes exist to score amino acid similarities.
- The sensitivity of these schemes in detecting distant relationships is not fully understood.
Purpose of the Study:
- To compare the effectiveness of four different protein sequence alignment methods.
- To determine which alignment approach is most sensitive for establishing evolutionary relationships.
- To assess performance across different levels of sequence identity.
Main Methods:
- Employed Needleman-Wunsch algorithm for similarity-based alignment.
- Utilized four weighting schemes: Unitary Matrix (UM), Genetic Code (GC), Structural/Genetic (SG), and Dayhoff Log-Odds Matrix (LOM).
- Applied methods to two protein families: globins and tyrosine kinase-like proteins.
Main Results:
- All methods agreed for sequences >30% identical.
- Significant alignment differences emerged for sequences <20% identical.
- Dayhoff LOM generally outperformed others in detecting distant relationships, validated by jumbling tests, but UM was sometimes equally effective.
Conclusions:
- Alignment method effectiveness varies with sequence identity.
- Dayhoff LOM is often superior for distant homology detection.
- Branching orders of phylogenetic trees were consistent across methods, though branch lengths varied.