def sailfish(fq1, fq2, sailfish_dir, gtf_file, ref_file, strandedness, data):
safe_makedir(sailfish_dir)
samplename = dd.get_sample_name(data)
quant_dir = os.path.join(sailfish_dir, "quant")
out_file = os.path.join(quant_dir, "quant.sf")
if file_exists(out_file):
return out_file
kmer_size = int(fastq.estimate_read_length(fq1))
if kmer_size < 30:
# kmer size must be odd
kmer_size = kmer_size - 5
kmer_size = kmer_size if kmer_size % 2 else kmer_size - 1
else:
kmer_size = 25
sailfish_idx = sailfish_index(gtf_file, ref_file, data, sailfish_dir, kmer_size)
num_cores = dd.get_num_cores(data)
sailfish = config_utils.get_program("sailfish", data["config"])
cmd = "{sailfish} quant -i {sailfish_idx} -p {num_cores} "
cmd += _libtype_string(fq1, fq2, strandedness)
fq1_cmd = "{fq1}" if not is_gzipped(fq1) else "<(gzip -cd {fq1})"
fq1_cmd = fq1_cmd.format(fq1=fq1)
if not fq2:
cmd += " -r {fq1_cmd} "
else:
fq2_cmd = "{fq2}" if not is_gzipped(fq2) else "<(gzip -cd {fq2})"
fq2_cmd = fq2_cmd.format(fq2=fq2)
cmd += " -1 {fq1_cmd} -2 {fq2_cmd} "
cmd += "--useVBOpt --numBootstraps 30 "
cmd += "-o {tx_out_dir}"
message = "Quantifying transcripts in {fq1} and {fq2}."
with file_transaction(data, quant_dir) as tx_out_dir:
do.run(cmd.format(**locals()), message.format(**locals()), None)
_sleuthify_sailfish(quant_dir)
return out_file
def counts_spikein(data):
data = utils.to_single_data(data)
samplename = dd.get_sample_name(data)
work_dir = dd.get_work_dir(data)
salmon_dir = os.path.join(work_dir, "spikein", samplename)
fasta_file = dd.get_spikein_fasta(data)
if not fasta_file:
return data
files = dd.get_input_sequence_files(data)
if len(files) == 2:
fq1, fq2 = files
else:
fq1, fq2 = files[0], None
assert file_exists(fasta_file), "%s was not found, exiting." % fasta_file
readlength = fastq.estimate_read_length(fq1)
if readlength % 2 == 0:
readlength -= 1
kmersize = min(readlength, 31)
logger.info("kmersize used for salmon index at spikein quant: %s" %
kmersize)
kmersize = kmersize if not dd.get_analysis(
data).lower() == "smallrna-seq" else 15
fasta_index = _index_spikein(fasta_file, salmon_dir, data, kmersize)
out_file = _salmon_quant_reads(fq1, fq2, salmon_dir, fasta_index, data)
data = dd.set_spikein_counts(data, out_file)
return data
def sailfish(fq1, fq2, sailfish_dir, gtf_file, ref_file, strandedness, data):
safe_makedir(sailfish_dir)
samplename = dd.get_sample_name(data)
quant_dir = os.path.join(sailfish_dir, "quant")
out_file = os.path.join(quant_dir, "quant.sf")
if file_exists(out_file):
return out_file
kmer_size = int(fastq.estimate_read_length(fq1))
if kmer_size < 30:
# kmer size must be odd
kmer_size = kmer_size - 5
kmer_size = kmer_size if kmer_size % 2 else kmer_size - 1
else:
kmer_size = 25
sailfish_idx = sailfish_index(gtf_file, ref_file, data, sailfish_dir,
kmer_size)
num_cores = dd.get_num_cores(data)
sailfish = config_utils.get_program("sailfish", data["config"])
cmd = "{sailfish} quant -i {sailfish_idx} -p {num_cores} "
cmd += _libtype_string(fq1, fq2, strandedness)
fq1_cmd = "{fq1}" if not is_gzipped(fq1) else "<(gzip -cd {fq1})"
fq1_cmd = fq1_cmd.format(fq1=fq1)
if not fq2:
cmd += " -r {fq1_cmd} "
else:
fq2_cmd = "{fq2}" if not is_gzipped(fq2) else "<(gzip -cd {fq2})"
fq2_cmd = fq2_cmd.format(fq2=fq2)
cmd += " -1 {fq1_cmd} -2 {fq2_cmd} "
cmd += "--useVBOpt --numBootstraps 30 "
cmd += "-o {tx_out_dir}"
message = "Quantifying transcripts in {fq1} and {fq2}."
with file_transaction(data, quant_dir) as tx_out_dir:
do.run(cmd.format(**locals()), message.format(**locals()), None)
_sleuthify_sailfish(quant_dir)
return out_file
def estimate_kmer_size(fq):
kmer_size = int(fastq.estimate_read_length(fq))
if kmer_size < 30:
# kmer size must be odd
kmer_size = kmer_size - 5
kmer_size = kmer_size if kmer_size % 2 else kmer_size - 1
else:
kmer_size = 25
return kmer_size
def estimate_kmer_size(fq):
kmer_size = int(fastq.estimate_read_length(fq))
if kmer_size < 30:
# kmer size must be odd
kmer_size = kmer_size - 5
kmer_size = kmer_size if kmer_size % 2 else kmer_size - 1
else:
kmer_size = 25
return kmer_size
def pick_kmersize(fq):
"""
pick an appropriate kmer size based off of https://www.biostars.org/p/201474/
tl;dr version: pick 31 unless the reads are very small, if not then guess
that readlength / 2 is about right.
"""
readlength = fastq.estimate_read_length(fq)
halfread = int(round(readlength / 2))
if halfread >= 31:
kmersize = 31
else:
kmersize = halfread
if kmersize % 2 == 0:
kmersize += 1
return kmersize
def pick_kmersize(fq):
"""
pick an appropriate kmer size based off of https://www.biostars.org/p/201474/
tl;dr version: pick 31 unless the reads are very small, if not then guess
that readlength / 2 is about right.
"""
readlength = fastq.estimate_read_length(fq)
halfread = int(round(readlength / 2))
if halfread >= 31:
kmersize = 31
else:
kmersize = halfread
if kmersize % 2 == 0:
kmersize += 1
return kmersize
def counts_spikein(data):
data = utils.to_single_data(data)
samplename = dd.get_sample_name(data)
work_dir = dd.get_work_dir(data)
salmon_dir = os.path.join(work_dir, "spikein", samplename)
fasta_file = dd.get_spikein_fasta(data)
if not fasta_file:
return data
files = dd.get_input_sequence_files(data)
if len(files) == 2:
fq1, fq2 = files
else:
fq1, fq2 = files[0], None
assert file_exists(fasta_file), "%s was not found, exiting." % fasta_file
readlength = fastq.estimate_read_length(fq1)
if readlength % 2 == 0:
readlength -= 1
kmersize = min(readlength, 31)
logger.info("kmersize used for salmon index at spikein quant: %s" % kmersize)
kmersize = kmersize if not dd.get_analysis(data).lower() == "smallrna-seq" else 15
fasta_index = _index_spikein(fasta_file, salmon_dir, data, kmersize)
out_file = _salmon_quant_reads(fq1, fq2, salmon_dir, fasta_index, data)
data = dd.set_spikein_counts(data, out_file)
return data