def count_seqs_in_file (pth): """ Largely just for checking we're processing data right. """ all_seqs = mcda.read_seqs (pth) return len (all_seqs)
def get_names_of_seqs_in_file (pth): all_seqs = mcda.read_seqs (pth) return [s.name for s in all_seqs]
"there are duplicate sequence names in the control set"
overlap = exp_seq_names.intersection (cntrl_seq_names)
assert len (overlap) == 0, \
"some sequences appear in experimental and control sets: %s" % \
', '.join (overlap)
## Main:
# create required directories
snakemake.utils.makedirs (BUILD_DIR)
snakemake.utils.makedirs (RESULTS_DIR)
snakemake.utils.makedirs (SEQ_WORK_DIR)
snakemake.utils.makedirs (COMP_SEQ_WORK_DIR)
# copy & rename controls, sample first 100 & make non-discovery set
all_control_seqs = mcda.read_seqs (input.all_cntrl_seqs)
mcda.write_seqs (all_control_seqs[:MEME_SEQ_CNT],
output.disc_cntrl_seqs)
mcda.write_seqs (all_control_seqs[MEME_SEQ_CNT:],
output.nondisc_control_seqs)
# copy & rename experimental, take top 100 & make non-discovery set
# NOTE: assumes exp seqs are in order of effect
all_exp_seqs = mcda.read_seqs (input.all_exp_seqs)
mcda.write_seqs (all_exp_seqs[:MEME_SEQ_CNT],
output.disc_exp_seqs)
mcda.write_seqs (all_exp_seqs[MEME_SEQ_CNT:],
output.nondisc_exp_seqs)
# copy all comparative seqs across
shell ("cp {input.comp_seq_data_dir}/*.fasta {output.comp_seq_work_dir}")
