def get_all_metrics():
sequences = get_list('InitialProtein')
first_ids = []
second_ids = []
first_lengths = []
second_lengths = []
all_identity = []
all_gaps = []
for sequence1, sequence2 in it.combinations(sequences, 2):
filename = get_pairwise_filename(sequence1, sequence2, aligned=True)
alignment_metrics = get_metrics_from_filename(filename)
first_ids.append(sequence1.id)
second_ids.append(sequence2.id)
first_lengths.append(alignment_metrics['first_length'])
second_lengths.append(alignment_metrics['second_length'])
all_identity.append(alignment_metrics['percent_identity'])
all_gaps.append(alignment_metrics['gap_fraction'])
metrics_df = pd.DataFrame({
'id1': first_ids,
'id2': second_ids,
'length1': first_lengths,
'length2': second_lengths,
'identity': all_identity,
'gaps': all_gaps
})
metrics_df['status'] = 'okay'
isoforms = (metrics_df.identity == 1) & (metrics_df.gaps > 0)
duplicates = (metrics_df.identity == 1) & (metrics_df.gaps == 0)
metrics_df.loc[isoforms, 'status'] = 'isoform'
metrics_df.loc[duplicates, 'status'] = 'duplicate'
return metrics_df
def make_sequence_files():
print('Splitting protein sequences into pairwise files...')
sequences = get_list('InitialProtein')
for sequence1, sequence2 in it.combinations(sequences, 2):
pairwise_filename = get_pairwise_filename(sequence1, sequence2)
pairwise_path = os.path.join('data', 'pairwise', pairwise_filename)
with open(pairwise_path, 'w') as output_file:
SeqIO.write([sequence1, sequence2], output_file, 'fasta')
def process(self, input):
"""
统计专利的合作数据, 作为Y
year + univ + 和别人合作数
"""
college_dir = input['college_dir']
college_list = get_list(college_dir)
if not os.path.exists('data.patent_pair'):
os.system('cat ' + input['patent_dir'] +'/* > data.patent_pair')
patent_dict = {}
ipc_hot_dict = {}
for line in open('data.patent_pair'):
items = line.strip().split('\t')
info = dict(zip(self.title_list, items))
open_id = info['公开（公告）号']
patent_type = 'A'
if open_id.endswith('U'):
patent_type = 'U'
apply_year = info['申请日'][:4]
apply_person = info['申请人']
hit_college = [] # 可能多个
for college in college_list:
if college in apply_person:
hit_college.append(college)
#
for college in hit_college:
key = apply_year + '\t' + college
if key not in patent_dict:
patent_dict[key] = {'A': 0, 'U': 0}
patent_dict[key][patent_type] += 1
# ### 统计热门IPC类别
# if '2012'<= apply_year <= '2018':
# ipc = info['主分类号'][:3]
# if ipc not in ipc_hot_dict:
# ipc_hot_dict[ipc] = 0
# ipc_hot_dict[ipc] += 1
# ##
# ff = open('res_ipc_hot1.csv', 'w')
# for ipc in ipc_hot_dict:
# ff.write(ipc + '\t' + str(ipc_hot_dict[ipc]) + '\n')
# ff.close()
return {'yy_dict': patent_dict,
'college_list': college_list
}
def single_linkage_clustering(threshold, variable='both'):
sequences = get_list('InitialProtein')
sequence_ids = [sequence.id for sequence in sequences]
number_of_sequences = len(sequences)
metrics = pairwise.get_all_metrics()
if variable == 'both':
correct_identity = metrics.identity > threshold['identity']
correct_gaps = metrics.gaps < threshold['gaps']
metrics['cluster_variable'] = correct_identity & correct_gaps
elif variable == 'identity':
metrics['cluster_variable'] = metrics.identity > threshold
current_assignment = np.arange(number_of_sequences)
for _, row in metrics[metrics['cluster_variable']].iterrows():
index1 = sequence_ids.index(row.id1)
index2 = sequence_ids.index(row.id2)
assignment1 = current_assignment[index1]
assignment2 = current_assignment[index2]
current_assignment[current_assignment==assignment1] = assignment2
return current_assignment
def cluster_and_align(threshold):
directory_path = os.path.join('data', 'clusters')
clusters = single_linkage_clustering(threshold)
sequences = get_list('InitialProtein')
count = 0
for cluster in set(clusters):
indices = np.arange(len(clusters))[clusters==cluster]
if len(indices) > 1:
filename = 'cluster_%d.fasta' % count
unaligned_path = os.path.join(directory_path, filename)
cluster_sequences = []
for index in indices:
cluster_sequences.append(sequences[index])
with open(unaligned_path, 'w') as output_file:
SeqIO.write(cluster_sequences, output_file, 'fasta')
aligned_filename = 'cluster_%d__ALIGNED.fasta' % count
aligned_path = os.path.join(directory_path, aligned_filename)
alignment_command = 'mafft %s > %s' % (unaligned_path, aligned_path)
subprocess.call(alignment_command, shell=True)
count += 1
from tools import create_file, create_directory, get_list, delete_file, copy_file, save_info, change_dir
from game import game
print('Hello. File manager is ready to work.')
command = None
save_info('Programm started')
while command != 'exit':
command = input(
'Enter command or exit to end work. Enter help to list commands: ')
if command == 'list':
get_list()
elif command == '0':
name = input('Enter new path: ')
change_dir(name)
elif command == '1':
name = input('Enter file name: ')
if name == '':
print('File name is missing')
save_info('Error - File name is missing')
else:
create_file(name)
save_info(f'File {name} is created')
elif command == '2':
name = input('Enter folder name: ')
if name == '':
print('Folder name is missing')
def setUp(self):
self.nucleotide_records = get_list('InitialNucleotide')
self.nucleotide_dictionary = get_dictionary('InitialNucleotide')
self.protein_dictionary = get_dictionary('InitialProtein')