Choice of Sequences in Phylogenetic Analysis 

Selecting the right sequences is the first and most critical step in any phylogenetic analysis. If the sequences chosen are inappropriate, the entire phylogeny constructed from them can turn out to be unreliable or even invalid, regardless of how carefully the subsequent computational analysis is carried out. The choice of sequences is therefore not arbitrary; it is guided both by the specific biological question being investigated and by certain technical requirements relating to sequence alignment and rate of evolution. A sound understanding of these principles at the outset ensures that the resulting phylogenetic tree is both accurate and biologically meaningful.

1. Biological Motivation for Choosing Sequences

Two main situations:

1. Studying phylogeny of organisms
   • Free to choose any gene from the organisms.
   • Gene should be informative about the pattern of species divergence.
2. Studying evolution of a gene/gene family
   • Same gene sequence is collected from as many species as possible.
   • Aim is to trace evolutionary history of that gene itself.

2. Criteria for Selecting a Suitable Gene

• Gene should have sequences available from many species.
• Degree of variation should be neither too high nor too low (moderate variation is ideal).

(a) Gene Evolving Too Slowly

• Very little variation among sequences.
• Not enough information to construct phylogeny.
• Useful only for very ancient/deep divergences.

(b) Gene Evolving Too Rapidly

• Sequences become highly divergent.
• Difficult to align sequences reliably.
• Difficult to estimate evolutionary distances accurately (due to multiple substitutions at same site).
• Phylogenetic methods become unreliable with highly divergent sequences.

Key point: Ideal gene = moderate rate of evolution (balance between too conserved and too variable).

3. Role of Multiple Sequence Alignment (MSA)

• Phylogenetic programs require MSA as input.
• Alignment assumes: nucleotides/amino acids in the same column are homologous.
• If this assumption is wrong → entire phylogeny becomes invalid.
• Results are very sensitive to alignment changes → alignment must be accurate.

Selection of Regions Within Sequence

• Some regions are highly variable, prone to insertions/deletions (indels).
• These regions:
  • Carry weak/no phylogenetic signal.
  • Add noise and increase uncertainty.
• Solution: Remove such unreliable variable regions before analysis.

Manual Checking of Alignment

• Automated alignment tools cannot be fully trusted.
• Manual checking/correction of alignment before phylogenetic work is essential and pays off in accuracy.

4. Example: Mitochondrial Small Subunit rRNA Gene (Primates)

• Gene length: approximately 950 bases.
• Sequences are relatively conserved within the group studied.
• Many invariant sites (shaded black in alignment) → increases confidence in alignment.
• Several variable sites → provide useful phylogenetic information.
• A few highly variable regions could not be aligned reliably → removed before analysis.

5. Taxonomic Groups Used in the Example

Based on NCBI Taxonomy Browser classification. Major primate groups included:

1. Strepsirhini – lemurs and related species
2. Tarsii – tarsiers
3. Platyrrhini – New World monkeys
4. Catarrhini – Old World monkeys and apes

Additional taxa included:

• 2 tree shrew species (Order: Scandentia) – considered close relatives of primates.
• 2 rodent species – used as outgroups.

6. Significance of Outgroup Choice

• Outgroups are used to root the tree and determine direction of evolution.
• Rodents are more distant than tree shrews, but recent studies place rodents + tree shrews + primates in one major mammalian clade.
• Hence rodents are a valid and useful outgroup here.

7. Value of This Data Set

• Species are well-studied, so true relationships are largely known.
• Data set includes:
  • Easy/straightforward relationships – correctly resolved by methods.
  • Difficult relationships – not always resolved correctly.
• Useful to test strengths and limitations of phylogenetic methods.

(Quick Recap)

• Choice of sequence = guided by biological question (organism phylogeny vs gene evolution).
• Gene must have moderate evolutionary rate (not too conserved, not too divergent).
• Accurate alignment is essential; homology assumption must hold.
• Highly variable/unreliable regions should be removed before analysis.
• Manual checking of automated alignments is necessary.
• Primate mtDNA rRNA example shows practical application: conserved core + informative variable sites + correct outgroup selection = reliable phylogeny.