def remake_with_star(in_path, out_path, title, comments='* added'):
'''Parses one of David Jimenez-Morales's BBTM matrices. Adds a
* column, which is equal to the minimum entry in the matrix.
'''
# Retrieve the matrix as a dictionary
mat = matrices.parse(in_path)
# Find the minimum entry
min_entry = min(mat.values())
# Make keys (i, *) and (*, i) map to that minimum entry for all
# letters i in complete_bbtm_ordering (including (*, *))
for i in complete_bbtm_ordering:
mat.update({(i, '*'): min_entry})
mat.update({('*', i): min_entry})
# Write the matrix
write_matrix(mat, out_path, title=title, comments=comments)
'Q': 3.7e-2,
'G': 7.4e-2,
'H': 2.9e-2,
'I': 3.8e-2,
'L': 7.6e-2,
'K': 7.2e-2,
'M': 1.8e-2,
'F': 4.0e-2,
'P': 5.0e-2,
'S': 8.1e-2,
'T': 6.2e-2,
'W': 1.3e-2,
'Y': 3.3e-2,
'V': 6.8e-2})
<<<<<<< HEAD
pam1 = 10**-4 * matrices.parse('pam1.txt')
=======
def expected_changes(mat, pi):
'''Expected fraction of positions different from original, after update
with mat, given original frequency distribution pi.
Very different from expected number of TRANSITIONS.'''
return sum((1 - mat[i,i])*pi[i] for i in pi.keys())
def avg_rate(q, pi):
return -1 * sum(q[i,i] *pi[i] for i in pi.keys())
def parse(mat_filename):
'''Take filename of a file in the format in which matrices are presented
in "Patterns of Amino Acid Substitutions..." Jimenez-Morales, Jie Liang