diff --git a/crisprbact/predict.py b/crisprbact/predict.py index f0e09b8c99b57f5c2b31407070dbff2d46b8a1d3..0c395ebefe923bf5f408259311d96f355ea2d55b 100644 --- a/crisprbact/predict.py +++ b/crisprbact/predict.py @@ -12,6 +12,7 @@ with open_binary("crisprbact", "reg_coef.pkl") as handle: coef = np.load(handle, allow_pickle=True) bases = ["A", "T", "G", "C"] +GUIDE_LEN = 20 def encode(seq): @@ -40,7 +41,7 @@ def find_targets(seq): guide_start = start - start_min guide_end = end - 16 - 3 guide = seq_revcomp[guide_start:guide_end] - assert len(guide) == 20 + assert len(guide) == GUIDE_LEN pos_seq_start = L - guide_end pos_seq_stop = L - guide_start pos_seq_pam = pos_seq_start - 3 @@ -62,7 +63,7 @@ def get_strand_value(value): return strand_dict[str(value)] -def on_target_predict(seq, genome=None, seed_sizes=[8, 9, 10, 11, 12]): +def on_target_predict(seq, genome=None, seed_sizes=[8, 9, 10, 11, 12, GUIDE_LEN]): seq = seq.upper() # make uppercase seq = re.sub(r"\s", "", seq) # removes white space