def RunTalesfTask(options):
logger = create_logger(options.logFilepath)
logger("Beginning")
if options.revcomp:
forwardOnly = False
else:
forwardOnly = True
if options.ncbi != "NA":
logger("Retrieving NCBI sequence. This could take a while if this sequence hasn't been used recently and needs to be downloaded from NCBI.")
with Conditional(options.ncbi != "NA", CachedEntrezFile(logger, options.ncbi)) as maybe_entrez_file:
if options.ncbi != "NA":
# Validate downloaded sequence
check_fasta_pasta(maybe_entrez_file.file)
if options.ncbi != "NA":
seqFilename = maybe_entrez_file.filepath
elif options.genome:
seqFilename = GENOME_FILE % options.organism
elif options.promoterome:
seqFilename = PROMOTEROME_FILE % options.organism
else:
seqFilename = options.fasta
result = ScoreTalesfTask(seqFilename, options.rvdString, options.outputFilepath, options.logFilepath, forwardOnly, options.cupstream, options.cutoff, 4, options.organism if options.genome else "")
if(result == 1):
raise TaskError()
def RunPairedTalesfTask(options):
logger = create_logger(options.logFilepath)
logger("Beginning")
if options.genome:
seqFilename = GENOME_FILE % options.organism
elif options.promoterome:
seqFilename = PROMOTEROME_FILE % options.organism
else:
seqFilename = options.fasta
result = ScorePairedTalesfTask(seqFilename, options.rvdString, options.rvdString2, options.outputFilepath, options.logFilepath, options.cupstream, options.cutoff, options.min, options.max, 4, options.organism if options.genome else "")
if(result == 1):
raise TaskError()
def RunTalesfTask(options):
logger = create_logger(options.logFilepath)
logger("Beginning")
if options.revcomp:
forwardOnly = False
else:
forwardOnly = True
if options.genome:
seqFilename = GENOME_FILE % options.organism
elif options.promoterome:
seqFilename = PROMOTEROME_FILE % options.organism
else:
seqFilename = options.fasta
result = ScoreTalesfTask(seqFilename, options.rvdString, options.outputFilepath, options.logFilepath, forwardOnly, options.cupstream, options.cutoff, 4, options.organism if options.genome else "")
if(result == 1):
raise TaskError()
parser.add_option('-z', '--nodeid', dest='nodeID', type='int', default = '-1', help='Drupal node ID')
parser.add_option('-k', '--ipaddr', dest='ip_address', type='string', default = '', help='IP address of job submitter')
(options, args) = parser.parse_args()
options.rvdString = options.rvdString.strip().upper()
validate_options_handler(validateOptions, options)
if options.genome:
queue_name = 'talesf_genome'
elif options.promoterome:
queue_name = 'talesf_promoterome'
else:
queue_name = 'talesf_other'
if options.nodeID != -1:
if not celery_found:
raise TaskError("nodeID option was provided but Celery is not installed")
logger = create_logger(options.logFilepath)
logger("Your task has been queued and will be processed when a worker node becomes available")
from findRvdTAL import TalesfTask
#if run from drupal then it should be queued as a task
TalesfTask.apply_async(kwargs=vars(options), queue=queue_name)
else:
RunTalesfTask(options)
def RunFindTALTask(options):
logger = create_logger(options.logFilepath)
logger("Beginning")
if options.check_offtargets and options.offtargets_ncbi != "NA":
logger(
"Retrieving NCBI off-target sequence. This could take a while if this sequence hasn't been used recently and needs to be downloaded from NCBI."
)
with Conditional(
options.check_offtargets and options.offtargets_ncbi != "NA", CachedEntrezFile(logger, options.offtargets_ncbi)
) as maybe_entrez_file:
if options.check_offtargets:
if not tfcount_found:
raise TaskError("Non off-target counting worker attempted to process off-target counting task.")
if options.offtargets_ncbi != "NA":
logger("Finished retrieving NCBI off-target sequence.")
# Validate downloaded sequence
check_fasta_pasta(maybe_entrez_file.file)
for record in FastaIterator(maybe_entrez_file.file, alphabet=generic_dna):
if len(record.seq) > OFFTARGET_COUNTING_SIZE_LIMIT:
raise TaskError(
"Off-Target counting is only supported for NCBI records where all individual sequences are under %d megabases in size"
% (OFFTARGET_COUNTING_SIZE_LIMIT / 1000000)
)
offtarget_seq_filename = ""
if options.offtargets_fasta != "NA":
offtarget_seq_filename = options.offtargets_fasta
elif options.offtargets_ncbi != "NA":
offtarget_seq_filename = maybe_entrez_file.filepath
elif options.genome:
offtarget_seq_filename = GENOME_FILE % options.organism
elif options.promoterome:
offtarget_seq_filename = PROMOTEROME_FILE % options.organism
else:
offtarget_seq_filename = options.fasta
strong_binding_RVDs = {"A": "NI", "C": "HD", "G": "NN", "T": "NG"}
if options.gspec:
strong_binding_RVDs["G"] = "NH"
seq_file = open(options.fasta, "r")
if options.outpath == "NA":
output_filepath = options.outdir + options.job + options.outfile
else:
output_filepath = options.outpath
out = open(output_filepath, "w")
table_ignores = ["TAL1 length", "TAL2 length", "Spacer length"]
out.write("table_ignores:" + ",".join(table_ignores) + "\n")
strand_min = 15 if options.arraymin is None else options.arraymin
strand_max = 20 if options.arraymax is None else options.arraymax
spacer_min = 15 if options.min is None else options.min
spacer_max = 30 if options.max is None else options.max
u_bases = []
if options.cupstream != 1:
u_bases.append("T")
if options.cupstream != 0:
u_bases.append("C")
out.write(
"options_used:"
+ ", ".join(
[
"array_min = " + str(strand_min),
"array_max = " + str(strand_max),
"spacer_min = " + str(spacer_min),
"spacer_max = " + str(spacer_max),
"upstream_base = " + (" or ".join(u_bases)),
]
)
+ "\n"
)
offtarget_header = "\tOff-Target Counts" if options.check_offtargets else ""
out.write(
"Sequence Name\tCut Site\tTAL1 start\tTAL2 start\tTAL1 length\tTAL2 length\tSpacer length\tSpacer range\tTAL1 RVDs\tTAL2 RVDs\tPlus strand sequence\tUnique RE sites in spacer\t% RVDs HD or NN/NH"
+ offtarget_header
+ "\n"
)
binding_sites = []
for gene in FastaIterator(seq_file, alphabet=generic_dna):
sequence = str(gene.seq).upper()
site_entry_counts = {}
if options.filter == 1:
if options.filterbase > len(sequence):
logger("Skipped %s as the provided cut site was greater than the sequence length" % (gene.id))
continue
cut_site_positions = [options.filterbase]
else:
cut_site_positions = range(len(sequence))
logger("Scanning %s for binding sites" % (gene.id))
for i in cut_site_positions:
cut_site_potential_sites = []
for spacer_size in range(spacer_min, spacer_max + 1):
spacer_potential_sites = []
spacer_size_left = int(math.floor(float(spacer_size) / 2))
spacer_size_right = int(math.ceil(float(spacer_size) / 2))
if i < (strand_min + spacer_size_left + 1) or i > (
len(sequence) - (strand_min + spacer_size_right) - 1
):
continue
for u_base in u_bases:
if u_base == "T":
d_base = "A"
elif u_base == "C":
d_base = "G"
u_pos_search_start = i - (strand_max + spacer_size_left) - 1
if u_pos_search_start < 0:
u_pos_search_start = 0
u_pos_search_end = i - (strand_min + spacer_size_left)
d_pos_search_start = i + (strand_min + spacer_size_right)
d_pos_search_end = i + (strand_max + spacer_size_right) + 1
u_positions = []
u_pos = 0
while True:
u_pos = sequence.rfind(u_base, u_pos_search_start, u_pos_search_end)
if u_pos == -1:
break
else:
u_pos_search_end = u_pos
u_positions.append(u_pos)
d_positions = []
d_pos = 0
while True:
d_pos = sequence.find(d_base, d_pos_search_start, d_pos_search_end)
if d_pos == -1:
break
else:
d_pos_search_start = d_pos + 1
d_positions.append(d_pos)
break_out = False
for u_pos in reversed(u_positions):
for d_pos in reversed(d_positions):
# uses inclusive start, exclusive end
tal1_start = u_pos + 1
tal1_end = i - spacer_size_left
tal1_seq = sequence[tal1_start:tal1_end]
tal2_start = i + spacer_size_right
tal2_end = d_pos
tal2_seq = sequence[tal2_start:tal2_end]
if not (
(tal1_seq in site_entry_counts and tal2_seq in site_entry_counts[tal1_seq])
or (tal1_seq in site_entry_counts and tal1_seq in site_entry_counts[tal1_seq])
or (tal2_seq in site_entry_counts and tal1_seq in site_entry_counts[tal2_seq])
or (tal2_seq in site_entry_counts and tal2_seq in site_entry_counts[tal2_seq])
):
bad_site = False
cg_count = 0
tal1_rvd = []
for c in tal1_seq:
if c not in strong_binding_RVDs:
bad_site = True
break
if c == "C" or c == "G":
cg_count += 1
tal1_rvd.append(strong_binding_RVDs[c])
if bad_site:
continue
tal1_rvd = " ".join(tal1_rvd)
tal2_rvd = []
for c in reverseComplement(tal2_seq):
if c not in strong_binding_RVDs:
bad_site = True
break
if c == "C" or c == "G":
cg_count += 1
tal2_rvd.append(strong_binding_RVDs[c])
if bad_site:
continue
tal2_rvd = " ".join(tal2_rvd)
if options.filter == 0:
break_out = True
binding_site = BindingSite(
seq_id=gene.id,
cutsite=i,
seq1_start=tal1_start,
seq1_end=tal1_end,
seq1_seq=tal1_seq,
seq1_rvd=tal1_rvd,
spacer_start=tal1_end,
spacer_end=tal2_start,
spacer_seq=sequence[tal1_end:tal2_start],
seq2_start=tal2_start,
seq2_end=tal2_end,
seq2_seq=tal2_seq,
seq2_rvd=tal2_rvd,
upstream=u_base,
cg_percent=int(
round(float(cg_count) / (len(tal1_seq) + len(tal2_seq)), 2) * 100
),
)
findRESitesInSpacer(sequence, binding_site)
if binding_site.seq1_seq not in site_entry_counts:
site_entry_counts[binding_site.seq1_seq] = {}
if binding_site.seq2_seq not in site_entry_counts[tal1_seq]:
site_entry_counts[binding_site.seq1_seq][binding_site.seq2_seq] = []
site_entry_counts[binding_site.seq1_seq][binding_site.seq2_seq].append(binding_site)
spacer_potential_sites.append(binding_site)
if break_out:
break
if break_out:
break
if len(spacer_potential_sites) > 0:
if options.filter == 0:
cut_site_potential_sites.append(reduce(filterByTALSize, spacer_potential_sites))
else:
cut_site_potential_sites.extend(spacer_potential_sites)
if len(cut_site_potential_sites) > 0:
if options.filter == 0:
binding_sites.append(reduce(filterByTALSize, cut_site_potential_sites))
else:
binding_sites.extend(cut_site_potential_sites)
if options.streubel:
binding_sites[:] = list(ifilterfalse(filterStreubel, binding_sites))
if options.check_offtargets:
if len(binding_sites) > 0:
off_target_pairs = []
for i, binding_site in enumerate(binding_sites):
off_target_pairs.append([binding_site.seq1_rvd, binding_site.seq2_rvd])
off_target_counts = PairedTargetFinderCountTask(
offtarget_seq_filename,
options.logFilepath,
options.cupstream,
3.0,
spacer_min,
spacer_max,
off_target_pairs,
)
for i, binding_site in enumerate(binding_sites):
binding_site.offtarget_counts = off_target_counts[i]
for i, binding_site in enumerate(binding_sites):
output_items = [
str(binding_site.seq_id),
str(binding_site.cutsite),
str(binding_site.seq1_start),
str(binding_site.seq2_end - 1),
str(binding_site.seq1_end - binding_site.seq1_start),
str(binding_site.seq2_end - binding_site.seq2_start),
str(binding_site.spacer_end - binding_site.spacer_start),
str(binding_site.spacer_start) + "-" + str(binding_site.spacer_end - 1),
binding_site.seq1_rvd,
binding_site.seq2_rvd,
binding_site.upstream
+ " "
+ binding_site.seq1_seq
+ " "
+ binding_site.spacer_seq.lower()
+ " "
+ binding_site.seq2_seq
+ " "
+ ("A" if binding_site.upstream == "T" else "G"),
binding_site.re_sites,
str(binding_site.cg_percent),
]
if options.check_offtargets:
output_items.append(" ".join(str(binding_site.offtarget_counts[x]) for x in range(5)))
out.write("\t".join(output_items) + "\n")
out.close()
seq_file.close()
logger("Finished")
def RunFindTALOldTask(options):
logger = create_logger(options.logFilepath)
seq_file = open(options.fasta, 'r')
logger("Beginning")
#Set other parameters
if options.arraymin is None or options.arraymax is None:
half_site_size = range(15, 31)
else:
half_site_size = range(options.arraymin, options.arraymax + 1)
if options.min is None or options.max is None:
spacer_size = [15, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
else:
spacer_size = range(options.min, options.max + 1)
#Initialize half site data structures:
gene_binding_sites = {}
#Open and read FASTA sequence file
genes = []
for gene in FastaIterator(seq_file, alphabet=generic_dna):
genes.append(gene)
seq_file.close()
for gene in genes:
gene.seq = gene.seq.upper()
#Set up binding site counter for each gene
binding_site_count_genes = {}
#Scan each gene sequence:
for gene in genes: #Scan sequence based on above criteria:
logger("Scanning %s for binding sites" % (gene.id))
binding_site_count_genes[gene] = 0
sequence = gene.seq
#Check each position along the sequence for possible binding sites using all combinations of binding site lengths and spacer lengths
for size1 in half_site_size:
for spacer_len in spacer_size:
for size2 in half_site_size:
for sindex in range(1, len(sequence)-(size1+spacer_len+size2)):
#Check for T at -1 for each half_site (A on plus strand of second halfsite)
if ((options.cupstream != 1 and sequence[sindex-1] == 'T' and sequence[sindex+size1+spacer_len+size2] == 'A') or (options.cupstream != 0 and sequence[sindex-1] == 'C' and sequence[sindex+size1+spacer_len+size2] == 'G')) and len(set(DNA) | set(sequence[sindex:sindex+size1+spacer_len+size2])) ==4:
half_site1 = sequence[sindex:sindex+size1]
half_site2_plus = sequence[sindex+size1+spacer_len:sindex+size1+spacer_len+size2]
half_site2_minus = sequence[sindex+size1+spacer_len:sindex+size1+spacer_len+size2].reverse_complement()
spacer_seq = sequence[sindex+size1:sindex+size1+spacer_len]
Binding_site_flag = True
#Check for not T at 1 (A at 1 for second halfsite) on both sites
if Binding_site_flag==True and options.t1==True:
if sequence[sindex] != 'T' and sequence[sindex+size1+spacer_len+size2-1]!= 'A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check not A at 2 (T on plus for second halfsite) on both sites
if Binding_site_flag==True and options.a2==True:
if sequence[sindex+1] !='A' and sequence[sindex+size1+spacer_len+size2-2]!='T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require T at end of both sites so bound by NG (A on plus for second half site)
if Binding_site_flag==True and options.tn==True:
if sequence[sindex+size1-1] == 'T' and sequence[sindex+size1+spacer_len] =='A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require last position to not be G's
if Binding_site_flag==True and options.gn==True:
if sequence[sindex+size1-1] != 'G' and sequence[sindex+size1+spacer_len] != 'C':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check nucleotide composition of the binding site
if Binding_site_flag==True and options.comp==True:
A1 = half_site1.count('A')/float(len(half_site1))
C1 = half_site1.count('C')/float(len(half_site1))
G1 = half_site1.count('G')/float(len(half_site1))
T1 = half_site1.count('T')/float(len(half_site1))
A2 = half_site2_minus.count('A')/float(len(half_site2_minus))
C2 = half_site2_minus.count('C')/float(len(half_site2_minus))
G2 = half_site2_minus.count('G')/float(len(half_site2_minus))
T2 = half_site2_minus.count('T')/float(len(half_site2_minus))
if A1<=percent_comp_range_top['A'] and A1>=percent_comp_range_bottom['A'] and C1<=percent_comp_range_top['C'] and C1>=percent_comp_range_bottom['C'] and G1<=percent_comp_range_top['G'] and G1>=percent_comp_range_bottom['G'] and T1<=percent_comp_range_top['T'] and T1>=percent_comp_range_bottom['T']:
if A2<=percent_comp_range_top['A'] and A2>=percent_comp_range_bottom['A'] and C2<=percent_comp_range_top['C'] and C2>=percent_comp_range_bottom['C'] and G2<=percent_comp_range_top['G'] and G2>=percent_comp_range_bottom['G'] and T2<=percent_comp_range_top['T'] and T2>=percent_comp_range_bottom['T']:
Binding_site_flag=True
else:
Binding_site_flag=False
else:
Binding_site_flag=False
#Create a binding site if all enforced rules have been met
if Binding_site_flag==True:
binding_site = Binding_site(perfectTAL1 = 'none', perfectTAL2 = 'none', start1 = sindex, start2=sindex+size1+spacer_len+size2-1, seq1 = half_site1, seq2_plus=half_site2_plus, spacer=spacer_len, spacerseq=spacer_seq, upstream = sequence[sindex-1])
if gene not in gene_binding_sites.keys():
gene_binding_sites[gene] = {}
if sindex not in gene_binding_sites[gene].keys():
gene_binding_sites[gene][sindex] = []
gene_binding_sites[gene][sindex].append(binding_site)
binding_site_count_genes[gene] +=1
#Compute TALs for each gene, using "strong-binding" RVDs for each nucleotide (binds the nucleotide more than half the time and we have more than 10 observations)
logger('Designing best scoring perfect TALs for each potential site...')
strong_binding_RVDs = {'A':'NI', 'C':'HD', 'G':'NN', 'T':'NG'}
RVD_pairs = {} #dictionary of RVD pair counts indexed by gene, RVD1 and RVD2
for gene in gene_binding_sites.keys():
RVD_pairs[gene] = {}
for start in gene_binding_sites[gene].keys():
#Find the perfect RVD sequence from each potential plus strand start site
for binding_site in gene_binding_sites[gene][start]:
TAL_1 = []
TAL_2 = []
for bindex in range(0, len(binding_site.seq1)):
TAL_1.append(strong_binding_RVDs[binding_site.seq1[bindex]])
TAL_1 = ' '.join(TAL_1)
for bindex in range(0, len(binding_site.seq2_minus)):
TAL_2.append(strong_binding_RVDs[binding_site.seq2_minus[bindex]])
TAL_2 = ' '.join(TAL_2)
binding_site.perfectTAL1 = TAL_1
binding_site.perfectTAL2 = TAL_2
if TAL_1 not in RVD_pairs[gene].keys():
RVD_pairs[gene][TAL_1] = {}
if TAL_2 not in RVD_pairs[gene][TAL_1].keys():
RVD_pairs[gene][TAL_1][TAL_2] = 0
RVD_pairs[gene][TAL_1][TAL_2] += 1
#Find lists of binding sites in each gene with unique (for that gene) perfect RVD sequences
binding_sites_unique_plus_minus_pairs = {} #binding sites whose RVD sequences don't make up other binding sites in the gene
for gene in gene_binding_sites.keys():
binding_sites_unique_plus_minus_pairs[gene] = {}
for start in gene_binding_sites[gene].keys():
for binding_site in gene_binding_sites[gene][start]:
RVD1 = binding_site.perfectTAL1
RVD2 = binding_site.perfectTAL2
if (RVD1 not in RVD_pairs[gene].keys() or RVD1 not in RVD_pairs[gene][RVD1].keys() or RVD_pairs[gene][RVD1][RVD1] == 1) and (RVD1 not in RVD_pairs[gene].keys() or RVD2 not in RVD_pairs[gene][RVD1].keys() or RVD_pairs[gene][RVD1][RVD2] == 1) and (RVD2 not in RVD_pairs[gene].keys() or RVD1 not in RVD_pairs[gene][RVD2].keys() or RVD_pairs[gene][RVD2][RVD1] == 1) and (RVD2 not in RVD_pairs[gene].keys() or RVD2 not in RVD_pairs[gene][RVD2].keys() or RVD_pairs[gene][RVD2][RVD2] == 1):
if start not in binding_sites_unique_plus_minus_pairs[gene].keys():
binding_sites_unique_plus_minus_pairs[gene][start] = []
binding_sites_unique_plus_minus_pairs[gene][start].append(binding_site)
#Check binding sites for unique Restriction endonuclease sites within the spacer.
#Unique sites are those that occur once in the spacer and do not occur in the 250 bases on either side of the spacer.
logger('Searching for restriction enzymes sites within each spacer...')
for gene in binding_sites_unique_plus_minus_pairs.keys():
for start_site in binding_sites_unique_plus_minus_pairs[gene].keys():
for binding_site in binding_sites_unique_plus_minus_pairs[gene][start_site]:
enzymes_in_spacer = []
spacer_start = binding_site.start1 + len(binding_site.seq1)
spacer_end = binding_site.start1 + len(binding_site.seq1) + binding_site.spacer-1
#identify sequence to check around the spacer for unique-ness
if spacer_start >= 250:
seq_check_start = spacer_start - 250
else:
seq_check_start = 0
if len(gene.seq) - spacer_end >= 250:
seq_check_end = spacer_end + 250 + 1
else:
seq_check_end = len(gene.seq)
#For each enzyme check if it occurs once in the spacer:
for enzyme in NEB_RE_sites:
if len(NEB_RE_sites[enzyme]["compiled"].findall(str(binding_site.spacerseq))) == 1:
#If unique in spacer, check that it doesen't occur in the flanking sequence
if len(NEB_RE_sites[enzyme]["compiled"].findall(str(gene.seq[seq_check_start : seq_check_end]))) == 1:
enzymes_in_spacer.append(enzyme)
#Create a string listing the enzymes and their sequences that can printed in the output
enzyme_string = ' '.join(["%s:%s" % (enzyme, NEB_RE_sites[enzyme]["short"]) for enzyme in enzymes_in_spacer])
if len(enzyme_string) == 0:
enzyme_string = 'none'
#append enzyme string to binding site object
binding_site.re_sites = enzyme_string
#Print output results to file: binding sites
if options.outpath == 'NA':
filename = options.outdir + options.job + options.outfile
else:
filename = options.outpath
out = open(filename, 'w')
table_ignores = ["TAL1 length", "TAL2 length", "Spacer length"]
out.write("table_ignores:" + string.join(table_ignores, ",") + "\n")
u_bases = []
if options.cupstream != 1:
u_bases.append("T")
if options.cupstream != 0:
u_bases.append("C")
out.write("options_used:" + ', '.join([
"array_min = " + str(options.arraymin),
"array_max = " + str(options.arraymax),
"spacer_min = " + str(options.min),
"spacer_max = " + str(options.max),
"upstream_base = " + (" or ".join(u_bases)),
("No T at position 1" if options.t1 else ""),
("No A at position 1" if options.a2 else ""),
("Sites must end in a T" if options.tn else ""),
("Sites may not end in G/NN" if options.gn else ""),
("Base composition rules enforced" if options.comp else ""),
]) + "\n")
out.write('Sequence Name\tTAL1 start\tTAL2 start\tTAL1 length\tTAL2 length\tSpacer length\tSpacer range\tTAL1 RVDs\tTAL2 RVDs\tPlus strand sequence\tUnique_RE_sites_in_spacer\n')
if len(binding_sites_unique_plus_minus_pairs.keys()) > 0:
for gene in binding_sites_unique_plus_minus_pairs.keys():
for start_site in binding_sites_unique_plus_minus_pairs[gene].keys():
for binding_site in binding_sites_unique_plus_minus_pairs[gene][start_site]:
out.write(gene.id + '\t' + str(binding_site.start1) + '\t' + str(binding_site.start2) + '\t' + str(len(binding_site.seq1)) + '\t' + str(len(binding_site.seq2_plus)) + '\t' + str(binding_site.spacer) + '\t' + str(binding_site.start1+len(binding_site.seq1)) + '-' + str(binding_site.start1+len(binding_site.seq1) + binding_site.spacer-1) + '\t' + binding_site.perfectTAL1 + '\t' + binding_site.perfectTAL2 + '\t' + binding_site.upstream.upper() + " " + str(binding_site.seq1) + ' ' + str(binding_site.spacerseq).lower() + ' ' + str(binding_site.seq2_plus) + " " + ("A" if binding_site.upstream == "T" else "G") + '\t' + binding_site.re_sites + '\n')
out.close()
logger('Finished')
def RunFindSingleTALSiteTask(options):
logger = create_logger(options.logFilepath)
logger("Beginning")
if options.check_offtargets and options.offtargets_ncbi != "NA":
logger("Retrieving NCBI off-target sequence. This could take a while if this sequence hasn't been used recently and needs to be downloaded from NCBI.")
with Conditional(options.check_offtargets and options.offtargets_ncbi != "NA", CachedEntrezFile(logger, options.offtargets_ncbi)) as maybe_entrez_file:
if options.check_offtargets:
if not tfcount_found:
raise TaskError("Non off-target counting worker attempted to process off-target counting task.")
if options.offtargets_ncbi != "NA":
logger("Finished retrieving NCBI off-target sequence.")
# Validate downloaded sequence
check_fasta_pasta(maybe_entrez_file.file)
for record in FastaIterator(maybe_entrez_file.file, alphabet=generic_dna):
if len(record.seq) > OFFTARGET_COUNTING_SIZE_LIMIT:
raise TaskError("Off-Target counting is only supported for NCBI records where all individual sequences are under %d megabases in size" % (OFFTARGET_COUNTING_SIZE_LIMIT / 1000000))
offtarget_seq_filename = ""
if options.offtargets_fasta != "NA":
offtarget_seq_filename = options.offtargets_fasta
elif options.offtargets_ncbi != "NA":
offtarget_seq_filename = maybe_entrez_file.filepath
elif options.genome:
offtarget_seq_filename = GENOME_FILE % options.organism
elif options.promoterome:
offtarget_seq_filename = PROMOTEROME_FILE % options.organism
else:
offtarget_seq_filename = options.fasta
strong_binding_RVDs = {
'A':'NI',
'C':'HD',
'G':'NN',
'T':'NG'
}
if options.gspec:
strong_binding_RVDs['G'] = 'NH'
seq_file = open(options.fasta, 'r')
#Set other parameters
if options.arraymin is None or options.arraymax is None:
half_site_size = range(15, 31)
else:
half_site_size = range(options.arraymin, options.arraymax + 1)
#Initialize half site data structures:
gene_binding_sites = {}
#Open and read FASTA sequence file
genes = []
for gene in FastaIterator(seq_file, alphabet=generic_dna):
genes.append(gene)
seq_file.close()
for gene in genes:
gene.seq = gene.seq.upper()
#Scan each gene sequence:
for gene in genes: #Scan sequence based on above criteria:
logger("Scanning %s for binding sites" % (gene.id))
sequence = gene.seq
#Check each position along the sequence for possible binding sites using all combinations of binding site lengths and spacer lengths
for size1 in half_site_size:
for sindex in range(1, len(sequence)-size1):
#Check for T at -1
if ((options.cupstream != 1 and sequence[sindex-1] == 'T') or (options.cupstream != 0 and sequence[sindex-1] == 'C')) and len(set(DNA) | set(sequence[sindex:sindex+size1])) ==4:
half_site1 = sequence[sindex:sindex+size1]
Binding_site_flag = True
#Check for not T at 1
if Binding_site_flag==True and options.t1==True:
if sequence[sindex] != 'T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check not A at 2
if Binding_site_flag==True and options.a2==True:
if sequence[sindex+1] !='A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require T at end
if Binding_site_flag==True and options.tn==True:
if sequence[sindex+size1-1] == 'T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require last position to not be G's
if Binding_site_flag==True and options.gn==True:
if sequence[sindex+size1-1] != 'G':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check nucleotide composition of the binding site
if Binding_site_flag==True and options.comp==True:
A1 = half_site1.count('A')/float(len(half_site1))
C1 = half_site1.count('C')/float(len(half_site1))
G1 = half_site1.count('G')/float(len(half_site1))
T1 = half_site1.count('T')/float(len(half_site1))
if A1<=percent_comp_range_top['A'] and A1>=percent_comp_range_bottom['A'] and C1<=percent_comp_range_top['C'] and C1>=percent_comp_range_bottom['C'] and G1<=percent_comp_range_top['G'] and G1>=percent_comp_range_bottom['G'] and T1<=percent_comp_range_top['T'] and T1>=percent_comp_range_bottom['T']:
Binding_site_flag=True
else:
Binding_site_flag=False
#Create a binding site if all enforced rules have been met
if Binding_site_flag==True:
binding_site = Binding_site(perfectTAL1 = 'none', start1 = sindex, seq1 = half_site1, is_plus=True, upstream=sequence[sindex-1])
if gene not in gene_binding_sites.keys():
gene_binding_sites[gene] = {}
if sindex not in gene_binding_sites[gene].keys():
gene_binding_sites[gene][sindex] = []
gene_binding_sites[gene][sindex].append(binding_site)
if options.revcomp==True: #Search for binding sites on the reverse complement strand
for sindex in range(size1-1, len(sequence)-1):
#Check for T at -1 for each half_site (A on plus strand)
if ((options.cupstream != 1 and sequence[sindex+1] == 'A') or (options.cupstream != 0 and sequence[sindex+1] == 'G')) and len(set(DNA) | set(sequence[sindex-size1+1:sindex+1])) == 4:
half_site1 = sequence[sindex-size1+1:sindex+1]
Binding_site_flag = True
#Check for not T at 1 (A at 1 on plus strand)
if Binding_site_flag==True and options.t1==True:
if sequence[sindex] != 'A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check not A at 2 (T on plus strand)
if Binding_site_flag==True and options.a2==True:
if sequence[sindex-1] !='T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require T at end so bound by NG (A on plus)
if Binding_site_flag==True and options.tn==True:
if sequence[sindex-size1+1] =='A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require last position to not be G (C on plus)
if Binding_site_flag==True and options.gn==True:
if sequence[sindex-size1+1] != 'C':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check nucleotide composition of the binding site
if Binding_site_flag==True and options.comp==True:
A2 = half_site1.count('T')/float(len(half_site1))
C2 = half_site1.count('G')/float(len(half_site1))
G2 = half_site1.count('C')/float(len(half_site1))
T2 = half_site1.count('A')/float(len(half_site1))
if A2<=percent_comp_range_top['A'] and A2>=percent_comp_range_bottom['A'] and C2<=percent_comp_range_top['C'] and C2>=percent_comp_range_bottom['C'] and G2<=percent_comp_range_top['G'] and G2>=percent_comp_range_bottom['G'] and T2<=percent_comp_range_top['T'] and T2>=percent_comp_range_bottom['T']:
Binding_site_flag=True
else:
Binding_site_flag=False
#Create a binding site if all enforced rules have been met
if Binding_site_flag==True:
binding_site = Binding_site(perfectTAL1 = 'none', start1 = sindex, seq1 = half_site1, is_plus=False, upstream=sequence[sindex+1])
if gene not in gene_binding_sites.keys():
gene_binding_sites[gene] = {}
if sindex not in gene_binding_sites[gene].keys():
gene_binding_sites[gene][sindex] = []
gene_binding_sites[gene][sindex].append(binding_site)
#Compute TALs for each gene, using "strong-binding" RVDs for each nucleotide (binds the nucleotide more than half the time and we have more than 10 observations)
logger('Designing best scoring perfect TALs for each potential site...')
for gene in gene_binding_sites.keys():
for start in gene_binding_sites[gene].keys():
#Find the perfect RVD sequence from each potential plus strand start site
for binding_site in gene_binding_sites[gene][start]:
TAL_1 = []
if binding_site.is_plus:
for bindex in range(0, len(binding_site.seq1)):
TAL_1.append(strong_binding_RVDs[binding_site.seq1[bindex]])
TAL_1 = ' '.join(TAL_1)
else:
rev_comp_seq = binding_site.seq1.reverse_complement()
for bindex in range(0, len(rev_comp_seq)):
TAL_1.append(strong_binding_RVDs[rev_comp_seq[bindex]])
TAL_1 = ' '.join(TAL_1)
binding_site.perfectTAL1 = TAL_1
#Print output results to file: binding sites
#filename = 'upload/'+ options.job + '_TALEN_pairs_all.txt'
if options.outpath == 'NA':
filename = options.outdir + options.job + options.outfile
else:
filename = options.outpath
binding_sites = []
if len(gene_binding_sites.keys()) > 0:
for gene in sorted(gene_binding_sites.keys()):
for start_site in gene_binding_sites[gene].keys():
for binding_site in gene_binding_sites[gene][start_site]:
binding_site.gene_id = gene.id
binding_sites.append(binding_site)
if options.check_offtargets:
if len(binding_sites) > 0:
off_target_seqs = []
for i, binding_site in enumerate(binding_sites):
off_target_seqs.append(binding_site.perfectTAL1)
off_target_counts = TargetFinderCountTask(offtarget_seq_filename, options.logFilepath, options.cupstream, 3.0, off_target_seqs)
for i, binding_site in enumerate(binding_sites):
binding_site.offtarget_count = off_target_counts[i]
out = open(filename, 'w')
table_ignores = []
if not options.revcomp:
table_ignores.append("Plus strand sequence")
if len(table_ignores) > 0:
out.write("table_ignores:" + string.join(table_ignores, ",") + "\n")
u_bases = []
if options.cupstream != 1:
u_bases.append("T")
if options.cupstream != 0:
u_bases.append("C")
out.write("options_used:" + ', '.join([
"array_min = " + str(options.arraymin),
"array_max = " + str(options.arraymax),
"upstream_base = " + (" or ".join(u_bases)),
("No T at position 1" if options.t1 else ""),
("No A at position 1" if options.a2 else ""),
("Sites must end in a T" if options.tn else ""),
("Sites may not end in G/NN" if options.gn else ""),
("Base composition rules enforced" if options.comp else ""),
("Search reverse complement" if options.revcomp else ""),
]) + "\n")
offtarget_header = "\tOff-Target Counts" if options.check_offtargets else ""
out.write('Sequence Name\tTAL start\tTAL length\tRVD sequence\tStrand\tTarget sequence\tPlus strand sequence' + offtarget_header + '\n')
for i, binding_site in enumerate(binding_sites):
offtarget_string = ""
if options.check_offtargets:
offtarget_string = "\t%d" % binding_site.offtarget_count
if binding_site.is_plus:
out.write(binding_site.gene_id + '\t' + str(binding_site.start1) + '\t' + str(len(binding_site.seq1)) + '\t' + binding_site.perfectTAL1 + '\t' + 'Plus' + '\t' + binding_site.upstream + " " + str(binding_site.seq1) + '\t' + binding_site.upstream + " " + str(binding_site.seq1) + offtarget_string + '\n')
else:
out.write(binding_site.gene_id + '\t' + str(binding_site.start1) + '\t' + str(len(binding_site.seq1)) + '\t' + binding_site.perfectTAL1 + '\t' + 'Minus' + '\t' + ("T" if binding_site.upstream == "A" else "C") + " " + str(binding_site.seq1.reverse_complement()) + '\t' + str(binding_site.seq1) + " " + binding_site.upstream + offtarget_string + '\n')
out.close()
logger('Finished')
def RunFindSingleTALSiteTask(options):
logger = create_logger(options.logFilepath)
strong_binding_RVDs = {
'A':'NI',
'C':'HD',
'G':'NN',
'T':'NG'
}
if options.gspec:
strong_binding_RVDs['G'] = 'NH'
seq_file = open(options.fasta, 'r')
#Set other parameters
if options.arraymin is None or options.arraymax is None:
half_site_size = range(15, 31)
else:
half_site_size = range(options.arraymin, options.arraymax + 1)
#Initialize half site data structures:
gene_binding_sites = {}
#Open and read FASTA sequence file
genes = []
for gene in FastaIterator(seq_file, alphabet=generic_dna):
genes.append(gene)
seq_file.close()
for gene in genes:
gene.seq = gene.seq.upper()
#Scan each gene sequence:
for gene in genes: #Scan sequence based on above criteria:
logger("Scanning %s for binding sites" % (gene.id))
sequence = gene.seq
#Check each position along the sequence for possible binding sites using all combinations of binding site lengths and spacer lengths
for size1 in half_site_size:
for sindex in range(1, len(sequence)-size1):
#Check for T at -1
if ((options.cupstream != 1 and sequence[sindex-1] == 'T') or (options.cupstream != 0 and sequence[sindex-1] == 'C')) and len(set(DNA) | set(sequence[sindex:sindex+size1])) ==4:
half_site1 = sequence[sindex:sindex+size1]
Binding_site_flag = True
#Check for not T at 1
if Binding_site_flag==True and options.t1==True:
if sequence[sindex] != 'T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check not A at 2
if Binding_site_flag==True and options.a2==True:
if sequence[sindex+1] !='A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require T at end
if Binding_site_flag==True and options.tn==True:
if sequence[sindex+size1-1] == 'T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require last position to not be G's
if Binding_site_flag==True and options.gn==True:
if sequence[sindex+size1-1] != 'G':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check nucleotide composition of the binding site
if Binding_site_flag==True and options.comp==True:
A1 = half_site1.count('A')/float(len(half_site1))
C1 = half_site1.count('C')/float(len(half_site1))
G1 = half_site1.count('G')/float(len(half_site1))
T1 = half_site1.count('T')/float(len(half_site1))
if A1<=percent_comp_range_top['A'] and A1>=percent_comp_range_bottom['A'] and C1<=percent_comp_range_top['C'] and C1>=percent_comp_range_bottom['C'] and G1<=percent_comp_range_top['G'] and G1>=percent_comp_range_bottom['G'] and T1<=percent_comp_range_top['T'] and T1>=percent_comp_range_bottom['T']:
Binding_site_flag=True
else:
Binding_site_flag=False
#Create a binding site if all enforced rules have been met
if Binding_site_flag==True:
binding_site = Binding_site(perfectTAL1 = 'none', start1 = sindex, seq1 = half_site1, is_plus=True, upstream=sequence[sindex-1])
if gene not in gene_binding_sites.keys():
gene_binding_sites[gene] = {}
if sindex not in gene_binding_sites[gene].keys():
gene_binding_sites[gene][sindex] = []
gene_binding_sites[gene][sindex].append(binding_site)
if options.revcomp==True: #Search for binding sites on the reverse complement strand
for sindex in range(size1-1, len(sequence)-1):
#Check for T at -1 for each half_site (A on plus strand)
if ((options.cupstream != 1 and sequence[sindex+1] == 'A') or (options.cupstream != 0 and sequence[sindex+1] == 'G')) and len(set(DNA) | set(sequence[sindex-size1+1:sindex+1])) == 4:
half_site1 = sequence[sindex-size1+1:sindex+1]
Binding_site_flag = True
#Check for not T at 1 (A at 1 on plus strand)
if Binding_site_flag==True and options.t1==True:
if sequence[sindex] != 'A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check not A at 2 (T on plus strand)
if Binding_site_flag==True and options.a2==True:
if sequence[sindex-1] !='T':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require T at end so bound by NG (A on plus)
if Binding_site_flag==True and options.tn==True:
if sequence[sindex-size1+1] =='A':
Binding_site_flag=True
else:
Binding_site_flag=False
#Require last position to not be G (C on plus)
if Binding_site_flag==True and options.gn==True:
if sequence[sindex-size1+1] != 'C':
Binding_site_flag=True
else:
Binding_site_flag=False
#Check nucleotide composition of the binding site
if Binding_site_flag==True and options.comp==True:
A2 = half_site1.count('T')/float(len(half_site1))
C2 = half_site1.count('G')/float(len(half_site1))
G2 = half_site1.count('C')/float(len(half_site1))
T2 = half_site1.count('A')/float(len(half_site1))
if A2<=percent_comp_range_top['A'] and A2>=percent_comp_range_bottom['A'] and C2<=percent_comp_range_top['C'] and C2>=percent_comp_range_bottom['C'] and G2<=percent_comp_range_top['G'] and G2>=percent_comp_range_bottom['G'] and T2<=percent_comp_range_top['T'] and T2>=percent_comp_range_bottom['T']:
Binding_site_flag=True
else:
Binding_site_flag=False
#Create a binding site if all enforced rules have been met
if Binding_site_flag==True:
binding_site = Binding_site(perfectTAL1 = 'none', start1 = sindex, seq1 = half_site1, is_plus=False, upstream=sequence[sindex+1])
if gene not in gene_binding_sites.keys():
gene_binding_sites[gene] = {}
if sindex not in gene_binding_sites[gene].keys():
gene_binding_sites[gene][sindex] = []
gene_binding_sites[gene][sindex].append(binding_site)
#Compute TALs for each gene, using "strong-binding" RVDs for each nucleotide (binds the nucleotide more than half the time and we have more than 10 observations)
logger('Designing best scoring perfect TALs for each potential site...')
for gene in gene_binding_sites.keys():
for start in gene_binding_sites[gene].keys():
#Find the perfect RVD sequence from each potential plus strand start site
for binding_site in gene_binding_sites[gene][start]:
TAL_1 = []
if binding_site.is_plus:
for bindex in range(0, len(binding_site.seq1)):
TAL_1.append(strong_binding_RVDs[binding_site.seq1[bindex]])
TAL_1 = ' '.join(TAL_1)
else:
rev_comp_seq = binding_site.seq1.reverse_complement()
for bindex in range(0, len(rev_comp_seq)):
TAL_1.append(strong_binding_RVDs[rev_comp_seq[bindex]])
TAL_1 = ' '.join(TAL_1)
binding_site.perfectTAL1 = TAL_1
#Print output results to file: binding sites
#filename = 'upload/'+ options.job + '_TALEN_pairs_all.txt'
if options.outpath == 'NA':
filename = options.outdir + options.job + options.outfile
else:
filename = options.outpath
out = open(filename, 'w')
table_ignores = []
if not options.revcomp:
table_ignores.append("Plus strand sequence")
if len(table_ignores) > 0:
out.write("table_ignores:" + string.join(table_ignores, ",") + "\n")
u_bases = []
if options.cupstream != 1:
u_bases.append("T")
if options.cupstream != 0:
u_bases.append("C")
out.write("options_used:" + ', '.join([
"array_min = " + str(options.arraymin),
"array_max = " + str(options.arraymax),
"upstream_base = " + (" or ".join(u_bases)),
("No T at position 1" if options.t1 else ""),
("No A at position 1" if options.a2 else ""),
("Sites must end in a T" if options.tn else ""),
("Sites may not end in G/NN" if options.gn else ""),
("Base composition rules enforced" if options.comp else ""),
("Search reverse complement" if options.revcomp else ""),
]) + "\n")
out.write('Sequence Name\tTAL start\tTAL length\tRVD sequence\tStrand\tTarget sequence\tPlus strand sequence\n')
if len(gene_binding_sites.keys()) > 0:
for gene in sorted(gene_binding_sites.keys()):
for start_site in gene_binding_sites[gene].keys():
for binding_site in gene_binding_sites[gene][start_site]:
if binding_site.is_plus:
out.write(gene.id + '\t' + str(binding_site.start1) + '\t' + str(len(binding_site.seq1)) + '\t' + binding_site.perfectTAL1 + '\t' + 'Plus' + '\t' + binding_site.upstream + " " + str(binding_site.seq1) + '\t' + binding_site.upstream + " " + str(binding_site.seq1) + '\n')
else:
out.write(gene.id + '\t' + str(binding_site.start1) + '\t' + str(len(binding_site.seq1)) + '\t' + binding_site.perfectTAL1 + '\t' + 'Minus' + '\t' + ("T" if binding_site.upstream == "A" else "C") + " " + str(binding_site.seq1.reverse_complement()) + '\t' + str(binding_site.seq1) + " " + binding_site.upstream + '\n')
out.close()
logger('Finished')