def temp_forward_unique_check():
geno = read_unambiguous(filename=PATH + FORWARD)
# comment()
s_arr = read_byte_numpy(append_file_name('data/22.sa'))
inv_suff, lcp = kasai(geno, s_arr)
myd = {}
for num in range(len(s_arr)):
myd[s_arr[num]] = lcp[num]
#trues0 = list(get_uniques(lcp))
json_it(trues0, "c22_forward_uniques")
def _part_0(args=None, print=print):
try:
start = time.time()
# _____________________________________________
print('\n_____________________________________')
print('PART 0: READ ARGS AND GENOME/SA FILES')
print('_____________________________________\n')
# _____________________________________________
past = start
print('reading SA...\n')
# read suffix array from bytes to ints
# reading with numpy then converting to 1-D array much slower than array.array
# however, array cannot read files larger than ~3GB
s_array = read_byte_numpy(filename=args.SA)
print('SA read.\n')
past = get_time(past, print=print)
print('reading genome...\n')
# read with Reads instead
# ! genome has ambs
#genome = reads(filename=args.genome)
chrs, genome = chr_splits(filename=args.genome)
json_it(data=chrs,
filename=append_file_name(args.outfile + "json_chrs"))
print('genome read.\n')
past = get_time(past, print=print)
# TODO: change below line as necessary
# args.LCPfile = '../data/lcp_pickle'
return genome, past, s_array, start
except Exception as e:
raise
def naive_lcp_22():
s_arr = fm.read_byte_numpy(filename=fm.append_file_name('data/22.sa'))
lcp = test_kasai.naive_lcp(s_array=s_arr,T=simple_genome())
fm.json_it(data=lcp,filename=fm.append_file_name('output/naive_lcp_22'))
def mu_driver():
"""
similar function as driver.py, except include minimal uniques instead of finding 20-100 uniquemers
:return:
"""
try:
# gitignore()
print('reading original genome: ', end='')
chrs, geno = chr_splits(filename=PATH + ORIGINAL)
json_it(chrs, append_file_name("json_chrs"))
del chrs
print('done.\nreading original SA...: ', end='')
s_arr = read_byte_numpy(append_file_name('data/genome.sa'))
lcp1 = kasai(geno, s_arr)[1]
d1 = OrderedDict(mu(SA=s_arr, LCP=lcp1))
del lcp1
del s_arr
au = _part_2(genome_file_name=PATH + ORIGINAL)
print("au list: ", list(au))
# *************************
# (2) flipped
# *************************
print("performing flips: ")
geno2 = read_unambiguous(PATH + FLIPPED)
s_arr2 = read_byte_numpy(append_file_name('data/flippedGeno.sa'))
lcp2 = kasai(geno2, s_arr2)[1]
del geno2
mu_result = dict(compare(d=d1, SA=s_arr2, LCP=lcp2))
del lcp2
mu_result = OrderedDict(sort_mu(mu_result))
mu_result = OrderedDict(true_address_dict(mu_result, au))
json_it(mu_result, append_file_name(files['MU_RESULT']))
#contigs = list(find_contigs(d=old_mu_result_without_true_addresses, bot=20, top=100))
contigs = OrderedDict(
find_perfect_contigs(d=mu_result, bot=20, top=100))
json_it(contigs, append_file_name(files['PERFECT_CONTIGS']))
contigs = list(within_distance(d=contigs, distance=300))
json_it(contigs,
append_file_name(files['PERFECT_CONTIGS_WITH_DISTANCE']))
print("number of contigs: ", len(contigs))
print("done")
except Exception as e:
raise
def _part_1(genome, past, s_array, args=None, print=print):
try:
# check if args.LCPfile exists
# if it does, read the pickle file instead of calculating new lcp
inv_suff = []
# ___________________________________________
print('\n_____________________________________')
print('PART 1: COMPUTE LCP ARRAY')
print('_____________________________________\n')
# ____________________________________________
# if user has specified a LCP pickle file that already exists
if args.lcpfile and os.path.isfile(path=args.lcpfile):
print("uniques file exists: ")
#print(args.lcpfile, '\n')
# TODO: change this as necessary
# hopefully start_uniques will be pickled/jsom/msgpacked in the future
#lcp = unpickle_dict(filename=args.lcpfile)
lcp = unjson_it(args.lcpfile)
# find out what format the lcp was pickled
if type(lcp) == dict or type(lcp) == OrderedDict:
key = lcp.keys()[0]
value = lcp[key]
if key > value and (100 >= value >= 20):
print("old lcp pickle was in format sa:lcp")
lcp = deque(lcp.values())
elif key < value and (100 >= key >= 20):
print("old lcp pickle was in format lcp:sa")
lcp = deque(lcp.keys())
else:
print("not sure what's going on here for sa_lcp dict")
raise KeyboardInterrupt
s_array = deque(s_array)
elif type(lcp) == list:
print('LCP file read as list')
print("uniques unpacked\n")
past = get_time(past, print=print)
print("Computing Unique Start Lengths")
# combine sa and lcp to form a dict with keys: sa, values: unique_starts
# TODO: creating OrderedDict consumes too much memory
filename = append_file_name('json_lcp')
if args.outfile:
filename = args.outfile
if args.inverse:
inv_suff = unjson_it(args.inverse)
else:
inv_suff = inverse1(s_array=s_array)
else:
if args.inverse:
inv_suff = unjson_it(args.inverse)
inv_suff, lcp = kasai(genome=genome,
inv_suff=inv_suff,
s_array=s_array,
print=print)
past = time.time()
# convert suffix array (list) to suffix array (deque) for increased efficiency
print('Completed.')
# json it
filename = append_file_name('json_lcp')
if args.outfile:
filename = append_file_name(args.outfile + 'json_lcp')
print('json\'ing lcp array to %s', filename)
json_it(data=lcp, filename=filename)
print('LCP json\'ed!')
past = get_time(past, print=print)
return past, lcp, inv_suff
except Exception as e:
raise