def save_template(plate_id, template):
for (cell, row, col) in iterate96WP():
if not template[cell]['prey']:
continue
bait = prey = None
is_NC = is_PC = False
if template[cell]['prey'] == POS_CONTROL:
is_PC = True
elif template[cell]['prey'] == NEG_CONTROL:
is_NC = True
else:
bait = template[cell]['bait']
prey = template[cell]['prey']
plate_cell = PlateCell(row=row,
column=col,
bait=bait,
prey=prey,
is_NC=is_NC,
is_PC=is_PC,
plate_id=plate_id)
db.add(plate_cell)
db.commit()
def init_db(proteins_list):
db_reset()
# Import proteins list
mapping = {
'A': 'family',
'C': 'id',
'I': 'symbol',
'J': 'long_symbol',
'K': 'description'
}
df_proteins = read_excel_list(proteins_list, 'Proteins List', mapping)
df_proteins['id'] = df_proteins['id'].str.upper()
df_proteins['symbol'].fillna(df_proteins.id, inplace=True)
# Validate data
df_invalid_id = df_proteins[~df_proteins.id.str.contains(
"^{}$".format(PROTEIN_ID_REGEX), regex=True, na=False)]
assert_empty_df(
df_invalid_id,
"Some of the provided ids in the protein list are invalid")
df_duplicate_id = df_proteins[df_proteins.duplicated(subset='id',
keep=False)]
assert_empty_df(
df_duplicate_id,
"Some of the provided ids in the protein list are duplicated")
df_no_family = df_proteins[df_proteins.family.isnull()]
assert_empty_df(
df_no_family,
"Some of the proteins in the reference list don't have an assigned family"
)
# Insert proteins into DB
proteins = df_proteins.to_dict(orient='records')
db.execute(Protein.__table__.insert(), proteins)
# Import batch sheets
mapping = {
'A': '#',
'B': 'subfamily',
'C': 'id',
'D': 'symbol',
'E': 'nickname',
'J': 'cloned'
}
for sheet in get_batch_sheets(proteins_list.name):
import_batch_sheet(proteins_list.name, sheet, mapping, df_proteins.id)
db.commit()
def plates_first_batch(first_batch_list, outfolder):
batch_id = 1
metadata = OrderedDict()
metadata['Notes'] = 'Generated from the original batch 1 template'
metadata['Batch 1 file'] = ''
plates, base_template = parse_1st_batch_list(first_batch_list)
# Index ids for batch, by sybmol
proteins = get_batch_proteins(batch_id)
prots_by_symbol = {protein.symbol:protein for protein in proteins}
replace_prey_symbols(base_template, prots_by_symbol)
# Remove plates from db
db.query(Plate).filter(Plate.bait_batch_id==batch_id,
Plate.prey_batch_id==batch_id
).delete()
for id, plate in enumerate(plates):
bait_1 = prots_by_symbol.get(plate['Bait_1'], None)
bait_2 = prots_by_symbol.get(plate['Bait_2'], None)
template = add_baits(base_template, bait_1, bait_2)
# Plate 4 of the first batch was put turned 180 in the reader
if plate['filename'] == 'batch 4 SERK1 BAK1':
template = reverse(template)
plate_id = id + 1
plate_name = screen_plate_name(plate_id)
metadata['Batch 1 file'] = plate['filename']
metadata['Timeshift'] = plate['timepoint']
# Save plate to database
plate = Plate(id=plate_id, bait_batch_id=batch_id, prey_batch_id=batch_id)
db.add(plate)
db.commit()
save_template(plate_id, template)
# TODO: export plate from database?
export_1st_batch_plate(plate_name, template, metadata, outfolder)
def create_screen_plates(bait_batch_id, prey_batch_id, outfolder):
metadata = OrderedDict()
metadata['Plate name'] = ''
metadata['Plate type'] = 'Screen'
metadata['Timeshift'] = ''
metadata['Bait plate'] = ''
metadata['Prey plate'] = storage_prey_plate_name(prey_batch_id)
# Same prey plate for all screen plates
prey_prots = get_batch_proteins(prey_batch_id)
prey_plate = PreyStoragePlate(prey_prots)
prey_plate.name = storage_prey_plate_name(prey_batch_id)
# 2 bait proteins per plate
bait_prots = get_batch_proteins(bait_batch_id)
for i in range(0, len(bait_prots), ScreenPlate.capacity()):
plate = Plate(bait_batch_id=bait_batch_id, prey_batch_id=prey_batch_id)
db.add(plate)
db.flush()
baits = bait_prots[i:i + ScreenPlate.capacity()]
bait_plate_name = storage_bait_plate_name(bait_batch_id, BaitStoragePlate.bait_plate_index(i))
screen_plate = ScreenPlate(baits, bait_plate_name, BaitStoragePlate.bait_plate_offset(i), prey_plate)
screen_plate.name = screen_plate_name(plate.id)
metadata['Plate name'] = screen_plate.name
metadata['Bait plate'] = bait_plate_name
metadata['Prey plate'] = screen_plate.prey_plate_name
template_file = os.path.join(outfolder, '{}_b{:02d}_p{:02d}_template.xlsx'.format(
screen_plate.name, bait_batch_id, prey_batch_id))
export_screen_plate(template_file, screen_plate, metadata)
logging.info("Saved screen template to {}".format(template_file))
# Save to database
db.commit()
def import_batch_sheet(filename, sheet, mapping, expected_ids):
'''
Import a batch sheet validating the data and checking that the
included ids are part of the reference protein list
'''
logging.info("Importing batch sheet '%s' from %s" % (sheet, filename))
batch_id = filter_digits(sheet)
batch_name = sheet[len('batch'):].strip()
# Import batch proteins
df_batch = read_excel_list(filename, sheet, mapping)
df_batch['id'] = df_batch['id'].str.upper()
df_batch['symbol'].fillna(df_batch.id, inplace=True)
# TODO: remove those not cloned?
logging.debug("Batch contains %d successfully cloned proteins" %
len(df_batch))
# Validate data
required = ['subfamily', 'id', 'symbol', 'nickname']
df_incomplete = df_batch[df_batch[required].isnull().any(axis=1)]
assert_empty_df(
df_incomplete,
"Some of the proteins in sheet '{}' have incomplete information".
format(sheet))
df_invalid_id = df_batch[~df_batch.id.str.contains(
"^{}$".format(PROTEIN_ID_REGEX), regex=True, na=False)]
assert_empty_df(
df_invalid_id,
"Some of the provided ids in sheet '{}' are invalid".format(sheet),
['#', 'id'])
df_duplicate_id = df_batch[df_batch.duplicated(subset='id', keep=False)]
assert_empty_df(
df_duplicate_id,
"Some of the provided ids in sheet '{}' are duplicated".format(sheet),
['#', 'id'])
df_unexpected_id = df_batch[~df_batch.id.isin(expected_ids)]
assert_empty_df(
df_unexpected_id,
"Some of the provided ids in sheet '{}' are not in the reference protein list"
.format(sheet), ['#', 'id'])
# Update the protein info with id, nickname, subfamily and symbol
proteins = df_batch.to_dict(orient='records')
for protein in proteins:
db.query(Protein).\
filter(Protein.id == protein['id']).\
update({k: protein[k] for k in ('subfamily', 'symbol', 'nickname')})
db.commit()
# Import the batch info
# TODO: order number? reindex?
df_batch_proteins = df_batch.copy()[['id', 'nickname', 'cloned']]
df_batch_proteins.rename(columns={'id': 'protein_id'}, inplace=True)
df_batch_proteins['batch_name'] = batch_name
df_batch_proteins['batch_id'] = batch_id
df_batch_proteins['order'] = df_batch_proteins.index
batch_proteins = df_batch_proteins.to_dict(orient='records')
db.execute(BatchProtein.__table__.insert(), batch_proteins)