def it_reverses_ptm_locs():
pros_df = pd.DataFrame(
[
("nul", False, 0, ""),
("id1", False, 1, ""),
("id2", False, 2, "1;3"),
],
columns=["pro_id", "pro_is_decoy", "pro_i", "pro_ptm_locs"],
)
pro_seqs_df = pd.DataFrame(
[
(0, "."),
(1, "B"),
(1, "C"),
(2, "D"),
(2, "E"),
(2, "F"),
(2, "G"),
],
columns=["pro_i", "aa"],
)
decoys_df, decoy_seqs_df = _step_3_generate_decoys(
pros_df, pro_seqs_df, decoy_mode="reverse")
# remember that the orgiinal df has the "null" entry at iloc 0, and no
# decoy is created for that.
assert decoys_df.iloc[0].pro_ptm_locs == ""
assert decoys_df.iloc[1].pro_ptm_locs == "2;4"
def it_handles_no_decoy_mode():
decoys_df, decoy_seqs_df = _step_3_generate_decoys(pros_df,
pro_seqs_df,
decoy_mode=None)
assert isinstance(decoys_df, pd.DataFrame) and len(decoys_df) == 0
assert isinstance(decoy_seqs_df,
pd.DataFrame) and len(decoy_seqs_df) == 0
def positive_reverse():
decoys_df, decoy_seqs_df = _step_3_generate_decoys(
pros_df, pro_seqs_df, decoy_mode="reverse")
def it_sets_is_decoy_true():
assert decoys_df.pro_is_decoy.tolist() == [True, True]
def it_skips_nul():
assert decoys_df.pro_i.tolist() == [3, 4]
def it_names_the_decoy():
assert decoys_df.pro_id.tolist() == ["rev-id1", "rev-id2"]
def it_reverses():
assert decoy_seqs_df.pro_i.tolist() == [3, 3, 4, 4]
assert decoy_seqs_df.aa.tolist() == ["C", "B", "E", "D"]
zest()
def positive_shuffle():
# add a new longer protein entry we can really test the shuffle with
nonlocal pros_df, pro_seqs_df
new_pro = {
"pro_id": "id3",
"pro_is_decoy": False,
"pro_i": 3,
"pro_ptm_locs": "",
}
pros_df = pros_df.append(new_pro, ignore_index=True)
aas = sorted(list(aaseq.aa_code_df.aa)) # 27 chars long
tups = list(itertools.zip_longest([], aas, fillvalue=3))
new_pro_seqs = pd.DataFrame(tups, columns=["pro_i", "aa"])
pro_seqs_df = pro_seqs_df.append(new_pro_seqs).reset_index(drop=True)
decoys_df, decoy_seqs_df = _step_3_generate_decoys(
pros_df, pro_seqs_df, decoy_mode="shuffle")
def it_sets_is_decoy_true():
assert decoys_df.pro_is_decoy.tolist() == [True, True, True]
def it_skips_nul():
assert decoys_df.pro_i.tolist() == [4, 5, 6]
def it_names_the_decoy():
assert decoys_df.pro_id.tolist() == [
"shu-id1", "shu-id2", "shu-id3"
]
def it_shuffles():
assert decoy_seqs_df.pro_i.tolist() == [4, 4, 5, 5
] + [6] * len(aas)
assert decoy_seqs_df.aa.tolist()[4:] != aas
assert decoy_seqs_df.aa.tolist()[4:] != aas[::-1]
assert sorted(
decoy_seqs_df.aa.tolist()[0:4]) == ["B", "C", "D", "E"]
assert sorted(decoy_seqs_df.aa.tolist()[4:]) == aas
zest()