Sequence length variation, indel costs, and congruence in sensitivity analysis

Area of Science:

• Bioinformatics
• Computational Biology
• Phylogenetics

Background:

• Sequence alignment is crucial for phylogenetic analysis.
• Indels (insertions/deletions) present significant challenges in sequence alignment.
• Evaluating congruence measures helps assess alignment quality.

Purpose of the Study:

• To investigate the performance of topological and character-based congruence measures.
• To assess the impact of different indel treatments on congruence.
• To identify optimal indel cost parameters for sequence alignment.

Main Methods:

• Analyzed two topological and four character-based congruence measures.
• Employed direct optimization within a sensitivity analysis framework.
• Varied indel costs, treating indels as a fifth character state or using linear affine gap costs.

Main Results:

• Congruence consistently improved when indels were treated as single events.
• No single congruence measure was consistently superior.
• All measures converged on the same optimal alignment cost set with sufficient data.
• Data partition dominance, characterized by length variation, caused initial disagreements.

Conclusions:

• Treating indels as single events enhances congruence in sequence alignment.
• Sufficient data integration overcomes disagreements among congruence measures.
• Data partition dominance can be detected and managed by adjusting indel costs.