def count_kmers_to_df(self, path_kmers):
""" Take all splits, count the kmer distribution and save to the kmer folder as pandas DataFrame """
for_csv = []
for segment, taxon, cat, start, end in self.yield_genome_split():
if cython_is_there:
kmer_count = cyt_ext.kmer_counter(str(segment.seq),
k=self.k,
dictionary=True,
combine=True)
else:
kmer_count = combine_counts_forward_w_rc(seq_count_kmer(
str(segment.seq),
deepcopy(self.kmer_count_zeros),
k=self.k),
k=self.k)
for_csv.append(
(taxon, cat, start, end, segment.name, segment.description,
self.path_fna, *kmer_count.values()))
# kmer_keys = list(self.kmer_count_zeros.keys())
df = pd.DataFrame(for_csv, columns=main.cols_types)
df.taxon = df.taxon.astype('category')
df.category = df.category.astype('category')
df.name = df.name.astype('category')
df.fna_path = df.fna_path.astype('category')
for col in self.col_kmers:
df[col] = df[col].astype("uint16") # todo: not using float32 ??
df.to_pickle(path_kmers)
logger.debug(f"saved kmer count to {path_kmers}")
def kmer_count(self, ignore_N=True):
""" common method """
if self._kmer_count is None:
if cython_is_there:
self._kmer_count = cyt_ext.kmer_counter(str(self.seq), k=K, dictionary=True, combine=True, length=len(self.seq))
else:
self._kmer_count = combine_counts_forward_w_rc(seq_count_kmer(self.seq, self.KMER.copy(), K, ignore_N=ignore_N), k=K)
return self._kmer_count
def count_kmers_to_df(self, path_kmers):
""" Take all splits, count the kmer distribution and save to the kmer folder as pandas DataFrame """
# todo: consider combining forward and backward kmers as well as complements.
# Single counter for AAAT, TAAA, TTTA and ATTT
for_csv = []
for segment, taxon, cat, start, end in self.yield_genome_split():
kmer_count = seq_count_kmer(str(segment.seq),
deepcopy(self.kmer_count_zeros),
k=self.k)
for_csv.append(
(taxon, cat, start, end, segment.name, segment.description,
self.path_fna, *kmer_count.values()))
# kmer_keys = list(self.kmer_count_zeros.keys())
df = pd.DataFrame(for_csv, columns=main.cols_types)
df.taxon = df.taxon.astype('category')
df.category = df.category.astype('category')
df.name = df.name.astype('category')
df.fna_path = df.fna_path.astype('category')
for col in self.col_kmers:
df[col] = df[col].astype("uint16")
df.to_pickle(path_kmers)
logger.debug(f"saved kmer count to {path_kmers}")
def test_bio_count_and_combine_seq(k, seq, counts):
assert bio.combine_counts_forward_w_rc(bio.seq_count_kmer(seq, k=k),
k) == counts
def test_bio_seq_count_kmer(k, seq, counts, np_counts):
assert bio.seq_count_kmer(seq, k=k) == counts