import genomeview
dataset_paths = [
"data/pacbio.chr1.bam", "data/illumina.chr1.bam",
"/Users/nspies/Downloads/hg19.refseq.sorted.bed.gz"
]
reference = "ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz"
chrom = "chr1"
start = 224368899
end = 224398899
doc = genomeview.visualize_data(dataset_paths, chrom, start, end, reference)
genomeview.save(doc, "example.svg")
# import genomeview.track
# chrom = "14"
# start = 66901400
# genome_path = "data/chr14.fa"
# doc = genomeview.Document(950)
# source = genomeview.FastaGenomeSource(genome_path)
# gv = genomeview.genomeview.GenomeView("bam", chrom, start, start+10000, "+", source)
# # Add the coordinate axis at the top
# axis = genomeview.axis.Axis("axis")
# gv.add_track(axis)
def build_context_image(hit_row,
alignment,
upstream_range=4000,
downstream_range=4000):
#extract hit accession of target from hit_row
hit_accession = hit_row.target_name
#extract start nt of target, stop nt of target and strand from hit_row
start = hit_row.seq_from
stop = hit_row.seq_to
seq_length = stop - start + 1
strand = hit_row.strand
#if strand is + then flip is true, if strand is - then flip is false. This is needed in the genome browser url.
flip_val = 'false'
if (strand == "-"):
flip_val = 'true'
#get the assembly accession of the hit and download the fasta and gff files
if hit_row.assembly_accession == 'nan':
base_filename = download_gff_fna(hit_accession)
else:
base_filename = hit_row.assembly_accession
fna_file = "data/raw/download/" + base_filename + "_genomic.fna"
gff_file = "data/raw/download/" + base_filename + "_genomic.gff"
#extract sequence of the part of the target that matches the query
target_sequence = find_seq_in_alignment(hit_row.target_coords, alignment)
target_sequence = target_sequence.replace('T', 'U')
#extract E-value
e_value = hit_row.e_value
#extract %gc
percent_gc = hit_row.gc
#extract score
score = hit_row.score
#extract target-name
target_name = hit_row.target_coords
#extract taxonomy
if hit_row.lineage == 'nan':
lineage = get_taxonomy(hit_accession)
else:
lineage = hit_row.lineage
#set range
down_limit = start - downstream_range
up_limit = stop + upstream_range
#print statements for variables to be shown in image
print("Match #{}".format(int(hit_row.name) + 1))
print("E-value: " + str(e_value))
print("%GC: " + str(percent_gc))
print("Score: " + str(score))
print("Genome Assembly: " + str(base_filename))
print("Target: " + target_name)
print("Lineage: " + lineage)
print("Matched Sequence: " + target_sequence)
#clickable link
genome_browser_url = 'https://www.ncbi.nlm.nih.gov/projects/sviewer/?id={}&v={}:{}&c=FF6600&theme=Details&flip={}&select=null&content=3&color=0&label=1&geneModel=0&decor=0&layout=0&spacing=0&alncolor=on&m={},{}&mn=5,3'.format(
hit_accession, down_limit, up_limit, flip_val, start, stop)
try:
display(
HTML(
'<a href="{}")>genome browser</a>'.format(genome_browser_url)))
except HTTPError as e:
if (e.code == 429):
time.sleep(0.5)
return display(
HTML('<a href="{}")>genome browser</a>'.format(
genome_browser_url)))
else:
raise
gff_file_zip = "/home/jovyan/work/data/raw/download/" + base_filename + "_genomic.gff.gz"
bed_file = "/home/jovyan/work/data/raw/features/" + base_filename + "_genomic.bed"
convert(gff_file_zip, bed_file, desc_only=True)
def prerender(renderer, element):
# prerenderers get run before the track is rendered
if start < stop:
x1 = element.scale.topixels(
start) # converting genomic coordinates to screen coordinates
x2 = element.scale.topixels(stop)
yield from renderer.rect(x1,
0,
x2 - x1,
element.height,
fill="lightblue",
stroke="none")
if start > stop:
x1 = element.scale.topixels(
start) # converting genomic coordinates to screen coordinates
x2 = element.scale.topixels(stop)
yield from renderer.rect(x1,
0,
x1 - x2,
element.height,
fill="lightblue",
stroke="none")
doc = genomeview.visualize_data({"": bed_file}, hit_accession, down_limit,
up_limit)
cur_track = genomeview.get_one_track(doc, "")
cur_track.prerenderers = [prerender]
display(doc)
return base_filename, lineage
def test_bed():
file_paths = ["data/genes.sorted.bed.gz"]
doc = genomeview.visualize_data(file_paths, "chr3", 179500230, 179800230)
genomeview.save(doc, "results/bed_view.svg")
def test_bed2(which_bed):
file_paths = ["data/{}.bed".format(which_bed)]
doc = genomeview.visualize_data(file_paths, "chr3", 179500230, 179800230)
genomeview.save(doc, "results/{}_view.svg".format(which_bed))
def build_context_image(hit_row,
alignment,
upstream_range=4000,
downstream_range=4000):
#extract hit accession of target from hit_row
hit_accession = hit_row.target_name
#extract start nt of target, stop nt of target and strand from hit_row
start = hit_row.seq_from
stop = hit_row.seq_to
seq_length = stop - start + 1
strand = hit_row.strand
#if strand is + then flip is true, if strand is - then flip is false. This is needed in the genome browser url.
flip_val = 'false'
if (strand == "-"):
flip_val = 'true'
#get the assembly accession of the hit and download the fasta and gff files
if hit_row.assembly_accession == 'nan':
base_filename = download_gff(hit_row)
else:
base_filename = hit_row.assembly_accession
fna_file = "data/raw/download/" + base_filename + "_genomic.fna"
gff_file = "data/raw/download/" + base_filename + "_genomic.gff"
#extract sequence of the part of the target that matches the query
target_sequence = find_seq_in_alignment(hit_row.target_coords, alignment)
target_sequence = target_sequence.replace('T', 'U')
#extract E-value
e_value = hit_row.e_value
#extract %gc
percent_gc = hit_row.gc
#extract score
score = hit_row.score
#extract target-name
target_name = hit_row.target_coords
#extract taxonomy
if hit_row.lineage == 'nan':
lineage = get_taxonomy(hit_row.tax_id)
else:
lineage = hit_row.lineage
#set range
down_limit = start - downstream_range
up_limit = stop + upstream_range
#don't let sequence ruler go negative
if down_limit < 0:
down_limit = 0
#force ruler to go slightly negative to make room for left "HIT" arrow
if flip_val and stop < 80:
down_limit = -70
#print statements for variables to be shown in image
print("Match #{}".format(int(hit_row.name) + 1))
print("E-value: " + str(e_value))
print("%GC: " + str(percent_gc))
print("Score: " + str(score))
print("Target: " + target_name)
print("Genome Assembly: " + str(base_filename))
print("Lineage: " + lineage)
print("Matched Sequence: " + target_sequence)
#skip browser link and context graph when there is no bed file found
if 'NO FEATURES' in base_filename or 'NOT FOUND' in base_filename or 'ERROR' in base_filename:
return ('nan', 'nan')
#clickable link
genome_browser_url = 'https://www.ncbi.nlm.nih.gov/projects/sviewer/?id={}&v={}:{}&c=FF6600&theme=Details&flip={}&select=null&content=3&color=0&label=1&geneModel=0&decor=0&layout=0&spacing=0&alncolor=on&m={},{}&mn=5,3'.format(
hit_accession, down_limit, up_limit, flip_val, start, stop)
try:
display(
HTML(
'<a href="{}")>genome browser</a>'.format(genome_browser_url)))
except HTTPError as e:
if (e.code == 429):
time.sleep(0.5)
return display(
HTML('<a href="{}")>genome browser</a>'.format(
genome_browser_url)))
else:
raise
gff_file_zip = "/home/jovyan/work/data/raw/download/" + base_filename + "_genomic.gff.gz"
bed_file = "/home/jovyan/work/data/raw/features/" + base_filename + "_genomic.bed"
convert(gff_file_zip, bed_file, desc_only=True)
def prerender(renderer, element):
# Prerenderers get run before the track is rendered.
# Draw non-feature graphic elements.
# IGR is in forward orientation
if start < stop:
x1 = element.scale.topixels(
start) # converting genomic coordinates to screen coordinates
x2 = element.scale.topixels(stop)
# Draw vertical hit bar
yield from renderer.rect(x1,
0,
x2 - x1,
element.height - 14,
fill="lightblue",
stroke="none")
# Add "HIT" text
yield from renderer.text(x1 + (x2 - x1) / 2,
element.height - 2,
"HIT",
size=12,
anchor="middle")
# Draw hit direction arrow (forward/positive stand)
yield from renderer.block_arrow(x1 + (x2 - x1) / 2 + 16,
element.height - 11,
9,
9,
arrow_width=9,
direction="right",
fill="black",
stroke="none")
# IGR is in reverse orientation
if start > stop:
x1 = element.scale.topixels(
start) # converting genomic coordinates to screen coordinates
x2 = element.scale.topixels(stop)
# Draw vertical hit bar
yield from renderer.rect(x2,
0,
x1 - x2,
element.height - 14,
fill="lightblue",
stroke="none")
# Add "HIT" text
yield from renderer.text(x1 + (x2 - x1) / 2,
element.height - 2,
"HIT",
size=12,
anchor="middle")
# Draw hit direction arrow (reverse/negative stand)
yield from renderer.block_arrow(x2 + (x1 - x2) / 2 - 25,
element.height - 11,
9,
9,
arrow_width=9,
direction="left",
fill="black",
stroke="none")
doc = genomeview.visualize_data({"": bed_file}, hit_accession, down_limit,
up_limit)
cur_track = genomeview.get_one_track(doc, "")
cur_track.prerenderers = [prerender]
display(doc)
return base_filename, lineage
def test_arrows(reference_path):
for i in [100, 1000, 2000, 5000]:
print(i)
doc = genomeview.visualize_data(["data/illumina.bam"], "chr4", 96549060, 96549060+i, reference_path)
genomeview.save(doc, "results/temp_{}.png".format(i))