def convert_to_xlsx(args):
"""
covert the .tsv file to xlsx
"""
samples_df = pd.read_csv(args.samples, comment="#", sep="\t")
column_names = [name.capitalize() for name in list(samples_df.columns)]
samples_sheet = []
for row in samples_df.itertuples(index=False, name='Pandas'):
method = getattr(row, "method")
condition = getattr(row, "condition")
replicate = getattr(row, "replicate")
fastq = getattr(row, "fastqFile")
fastq = fastq.split("/")[1]
ntup = [method, condition, replicate, fastq]
if len(samples_df.columns) == 5:
fastq2 = getattr(row, "fastqFile2")
fastq2 = fastq2.split("/")[1]
ntup.append(fastq2)
samples_sheet.append(ntup)
samples_df = pd.DataFrame.from_records(samples_sheet, columns=column_names)
sheets = {"samples": samples_df}
eu.excel_writer(args.output, sheets, [])
def create_excel_sheets(args):
excel_sheet_dict = {}
# read the genome file
genome_file = SeqIO.parse(args.genome_path, "fasta")
genome_dict = dict()
for entry in genome_file:
genome_dict[str(entry.id)] = (str(entry.seq),
str(entry.seq.complement()))
# get the total mapped reads for each bam file
total_mapped_dict = {}
with open(args.total_mapped, "r") as f:
total = f.readlines()
wildcards = []
for line in total:
wildcard, reference_name, value = line.strip().split("\t")
total_mapped_dict[(wildcard, reference_name)] = int(value)
wildcards.append(wildcard)
wildcards = eu.get_unique(wildcards)
te_header = eu.get_te_header(wildcards)
conditions = []
for card in wildcards:
conditions.append(card.split("-")[1])
conditions = eu.get_unique(conditions)
contrasts = sorted(
[f"{x}-{y}" for x, y in list(iter.combinations(conditions, 2))])
excel_sheet_dict = create_cds_excel_sheet(args, excel_sheet_dict,
genome_dict, total_mapped_dict,
wildcards, conditions, contrasts,
te_header)
excel_sheet_dict = create_misc_excel_sheet(args, excel_sheet_dict,
genome_dict, total_mapped_dict,
wildcards, conditions,
contrasts, te_header)
eu.excel_writer(args.output_path, excel_sheet_dict, wildcards)
def create_excel_file(args):
# read the genome file
genome_file = SeqIO.parse(args.genome, "fasta")
genome_dict = dict()
for entry in genome_file:
genome_dict[str(entry.id)] = (str(entry.seq), str(entry.seq.complement()))
# get the total mapped reads for each bam file
total_mapped_dict = {}
with open(args.total_mapped, "r") as f:
total = f.readlines()
wildcards = []
for line in total:
wildcard, chromosome, value = line.strip().split("\t")
total_mapped_dict[(wildcard, chromosome)] = int(value)
wildcards.append(wildcard)
wildcards = eu.get_unique(wildcards)
te_header = eu.get_te_header(wildcards)
conditions = []
for card in wildcards:
conditions.append(card.split("-")[1])
conditions = eu.get_unique(conditions)
#read bed file
read_df = pd.read_csv(args.reads, comment="#", header=None, sep="\t")
# read gff file
cds_sheet = []
header = ["Identifier", "Genome", "Source", "Feature", "Start", "Stop", "Strand", "Pred_probability", "Locus_tag", "Old_locus_tag", "Name", "Length", "Codon_count"] + [cond + "_TE" for cond in te_header] + [card + "_rpkm" for card in wildcards] + ["Evidence", "Start_codon", "Stop_codon", "15nt upstream", "Nucleotide_seq", "Aminoacid_seq"]
prefix_columns = len(read_df.columns) - len(wildcards)
name_list = ["s%s" % str(x) for x in range(len(header))]
nTuple = collections.namedtuple('Pandas', name_list)
for row in read_df.itertuples(index=False, name='Pandas'):
chromosome = getattr(row, "_0")
source = getattr(row, "_1")
feature = getattr(row, "_2")
start = getattr(row, "_3")
stop = getattr(row, "_4")
strand = getattr(row, "_6")
attributes = getattr(row, "_8")
start_codon, stop_codon, nucleotide_seq, aa_seq, nt_window = eu.get_genome_information(genome_dict[chromosome], start-1, stop-1, strand)
pred_value, name, product, note, evidence, locus_tag, old_locus_tag = eu.retrieve_column_information(attributes)
length = stop - start + 1
codon_count = int(length / 3)
read_list = [getattr(row, "_%s" %x) for x in range(prefix_columns,len(row))]
rpkm_list = []
for idx, val in enumerate(read_list):
rpkm_list.append(eu.calculate_rpkm(total_mapped_dict[(wildcards[idx], chromosome)], val, length))
te_list = eu.calculate_te(rpkm_list, wildcards, conditions)
identifier = "%s:%s-%s:%s" % (chromosome, start, stop, strand)
result = [identifier, chromosome, "reparation", feature, start, stop, strand, pred_value, locus_tag, old_locus_tag, name, length, codon_count] + te_list + rpkm_list + [evidence, start_codon, stop_codon, nt_window, nucleotide_seq, aa_seq]
cds_sheet.append(nTuple(*result))
cds_df = pd.DataFrame.from_records(cds_sheet, columns=[header[x] for x in range(len(header))])
cds_df = cds_df.sort_values(by=["Genome", "Start", "Stop"])
dataframe_dict = { "CDS" : cds_df }
eu.excel_writer(args.output_path, dataframe_dict, wildcards)
def parse_orfs(args):
with open(args.total_mapped, "r") as f:
total = f.readlines()
wildcards = []
for line in total:
wildcard, reference_name, value = line.strip().split("\t")
wildcards.append(wildcard)
wildcards = get_unique(wildcards)
#read bed file
read_df = pd.read_csv(args.reads, comment="#", header=None, sep="\t")
# read gff file
main_sheet = []
header = ["Orientation", "Class", "Feature count"] + wildcards
prefix_columns = len(read_df.columns) - len(wildcards)
name_list = ["s%s" % str(x) for x in range(len(header))]
nTuple = collections.namedtuple('Pandas', name_list)
decode = {
"srna": "sRNA",
"5'-utr": "5'-UTR",
"cds": "CDS",
"rrna": "rRNA",
"trna": "tRNA",
"transcript": "transcript",
"pseudogene": "pseudogene",
"total": "total"
}
feature_list = [
"srna", "5'-utr", "cds", "rrna", "trna", "transcript", "pseudogene",
"total"
]
read_dict = collections.OrderedDict()
count_dict = collections.OrderedDict()
for f in feature_list:
read_dict[f] = [0] * len(wildcards)
count_dict[f] = 0
main_sheet = []
for row in read_df.itertuples(index=False, name='Pandas'):
reference_name = getattr(row, "_0")
start = getattr(row, "_3")
stop = getattr(row, "_4")
feature = getattr(row, "_2")
read_list = [
getattr(row, "_%s" % x) for x in range(prefix_columns, len(row))
]
if feature.lower() in read_dict:
for idx, value in enumerate(read_list):
read_dict[feature.lower()][idx] += value
read_dict["total"][idx] += value
count_dict[feature.lower()] += 1
count_dict["total"] += 1
else:
print("feature not usable: " + feature)
for key, val in read_dict.items():
result = ["sense", decode[key], count_dict[key]
] + [float("%.2f" % v) for v in val]
main_sheet.append(nTuple(*result))
main_df = pd.DataFrame.from_records(
main_sheet, columns=[header[x] for x in range(len(header))])
dataframe_dict = {"Main": main_df}
eu.excel_writer(args.output, dataframe_dict, wildcards)
def xtail_output(args):
# read the genome file
genome_file = SeqIO.parse(args.genome, "fasta")
genome_dict = dict()
for entry in genome_file:
genome_dict[str(entry.id)] = (str(entry.seq),
str(entry.seq.complement()))
annotation_dict = annotation_to_dict(args.annotation_file)
diff_expr_df = pd.read_csv(args.input_csv, sep=",", comment="#")
all_sheet = []
header = [
"Genome", "Start", "Stop", "Strand", "Locus_tag", "Old_locus_tag",
"ID", "Name", "mRNA_log2FC", "RPF_log2FC", "log2FC_TE_v1", "pvalue_v1",
"log2FC_TE_v2", "pvalue_v2", "log2FC_TE_final", "pvalue_final",
"pvalue.adjust", "Length", "Codon_count", "Start_codon", "Stop_codon",
"Nucleotide_seq", "Aminoacid_seq"
]
name_list = [f"s{x}" for x in range(len(header))]
nTuple = collections.namedtuple('Pandas', name_list)
for row in diff_expr_df.itertuples(index=False, name='Pandas'):
cds_id = getattr(row, "_0")
if cds_id in annotation_dict:
chromosome, start, stop, strand, attributes, locus_tag, old_locus_tag = annotation_dict[
cds_id]
column_info = retrieve_column_information(attributes)
else:
if ":" in cds_id and "-" in cds_id:
chromosome, sec, strand = cds_id.split(":")
start, stop = sec.split("-")
column_info = ["", "", ""]
locus_tag, old_locus_tag = "", ""
else:
sys.exit("Error... ID is not novel and not in the annotation!")
mRNA_log2FC = getattr(row, "mRNA_log2FC")
RPF_log2FC = getattr(row, "RPF_log2FC")
log2FC_TE_v1 = getattr(row, "log2FC_TE_v1")
pvalue_v1 = getattr(row, "pvalue_v1")
log2FC_TE_v2 = getattr(row, "log2FC_TE_v2")
pvalue_v2 = getattr(row, "pvalue_v2")
log2FC_TE_final = getattr(row, "log2FC_TE_final")
pvalue_final = getattr(row, "pvalue_final")
pvalue_adjust = getattr(row, "_9")
start = int(start)
stop = int(stop)
length = stop - start + 1
codon_count = int(length / 3)
start_codon, stop_codon, nucleotide_seq, aa_seq = get_genome_information(
genome_dict[chromosome], start - 1, stop - 1, strand)
result = [
chromosome, start, stop, strand, locus_tag, old_locus_tag, cds_id,
column_info[0], mRNA_log2FC, RPF_log2FC, log2FC_TE_v1, pvalue_v1,
log2FC_TE_v2, pvalue_v2, log2FC_TE_final, pvalue_final,
pvalue_adjust, length, codon_count, start_codon, stop_codon,
nucleotide_seq, aa_seq
]
all_sheet.append(nTuple(*result))
all_df = pd.DataFrame.from_records(
all_sheet, columns=[header[x] for x in range(len(header))])
dataframe_dict = {"all": all_df}
eu.excel_writer(args.output, dataframe_dict, [])