- Local Minimum Problem
- Parameter Choice Problem

1. Local Minimum Problem

- It stems from greedy nature of alignment
(mistakes made early in alignment cannot be corrected later)

- A better tree gives a better alignment
(UPGMA neighbour-joining tree method)
 

2. Parameter Choice Problem

- It stems from using just one set of parameters
(and hoping that they will do for all)

Parameters:

i. amino acid substitution matrix

ii. Gap Opening Penalty

iii. Gap Extension Penalty