def get_region_from_indiv(indiv,wssd_lib_dir,bac_analysis_lib_dir,chr,start,end,fn_gc_table,fn_contig):
#mask_table = DenseTrackSet(fn_contig,
# fn_mask_table,
# overwrite=False,
# openMode='r')
gc_content_table = DenseTrackSet(fn_contig,
fn_gc_table,
overwrite=False,
openMode='r')
read_depth = np.zeros(end-start)
print(read_depth)
print(end-start)
libls = os.listdir(bac_analysis_lib_dir)
for lib in libls:
if(lib[0:1]=="_"):
continue
lib_qc_dir = "%s/%s"%(bac_analysis_lib_dir,lib)
if(kgf.lib_pass_qc(lib_qc_dir)):
print("library %s passed qc: loading..."%(lib))
fn_wssd = "%s/%s/hg18rmsk.wssd"%(wssd_lib_dir,lib)
gc_correction_vect = load_gc_correction_vect(lib_qc_dir)
print("opening wssd %s"%(fn_wssd))
wssd = WssdFile(fn_contig,
fn_wssd,
overwrite=False,
openMode='r')
print("wssd opened..")
grpWssd = "wssd"
grpGC = "GC_content"
read_depth+=get_adjusted_depth(wssd,grpWssd,gc_content_table,grpGC,chr,start,end,gc_correction_vect,0)
wssd.tbl.close()
del wssd
gc_content_table.tbl.close()
del gc_content_table
return read_depth
def get_region_from_indiv(indiv,wssd_lib_dir,bac_analysis_lib_dir,chr,start,end,fn_gc_table,fn_contig):
#mask_table = DenseTrackSet(fn_contig,
# fn_mask_table,
# overwrite=False,
# openMode='r')
gc_content_table = DenseTrackSet(fn_contig,
fn_gc_table,
overwrite=False,
openMode='r')
read_depth = np.zeros(end-start)
print read_depth
print end-start
libls = os.listdir(bac_analysis_lib_dir)
for lib in libls:
if(lib[0:1]=="_"):
continue
lib_qc_dir = "%s/%s"%(bac_analysis_lib_dir,lib)
if(kgf.lib_pass_qc(lib_qc_dir)):
print "library %s passed qc: loading..."%(lib)
fn_wssd = "%s/%s/hg18rmsk.wssd"%(wssd_lib_dir,lib)
gc_correction_vect = load_gc_correction_vect(lib_qc_dir)
print "opening wssd %s"%(fn_wssd)
wssd = WssdFile(fn_contig,
fn_wssd,
overwrite=False,
openMode='r')
print "wssd opened.."
grpWssd = "wssd"
grpGC = "GC_content"
read_depth+=get_adjusted_depth(wssd,grpWssd,gc_content_table,grpGC,chr,start,end,gc_correction_vect,0)
wssd.tbl.close()
del wssd
gc_content_table.tbl.close()
del gc_content_table
return read_depth
def get_passing_libs(bac_analysis_lib_dir,max_correction,overide_thresh):
passing_libs = {}
libls = os.listdir(bac_analysis_lib_dir)
for lib in libls:
if(lib[0:1]=="_" or lib[0] == "."):
continue
lib_qc_dir = "%s/%s"%(bac_analysis_lib_dir,lib)
if(kgf.lib_pass_qc(lib_qc_dir,overide_thresh)):
print "lib passed: %s"%(lib)
gc_correction_vects = []
#gc_correction_vects.append(load_gc_correction_vect(lib_qc_dir,0))
#gc_correction_vects.append(load_gc_correction_vect(lib_qc_dir,1))
gc_correction_vects.append(kgf.get_GC_depth_correction_from_dir(lib_qc_dir,max_correction))
#print gc_correction_vects[0]
passing_libs[lib] = gc_correction_vects
return passing_libs
def get_passing_libs(bac_analysis_lib_dir, max_correction, overide_thresh):
passing_libs = {}
libls = os.listdir(bac_analysis_lib_dir)
for lib in libls:
if (lib[0:1] == "_" or lib[0] == "."):
continue
lib_qc_dir = "%s/%s" % (bac_analysis_lib_dir, lib)
if (kgf.lib_pass_qc(lib_qc_dir, overide_thresh)):
print "lib passed: %s" % (lib)
gc_correction_vects = []
#gc_correction_vects.append(load_gc_correction_vect(lib_qc_dir,0))
#gc_correction_vects.append(load_gc_correction_vect(lib_qc_dir,1))
gc_correction_vects.append(
kgf.get_GC_depth_correction_from_dir(lib_qc_dir,
max_correction))
#print gc_correction_vects[0]
passing_libs[lib] = gc_correction_vects
return passing_libs
def analyze_coverage_to_correlation(analysis_dir):
#anal_by_alg = {}
print "************************"
print "analyze_coverage_to_correlation"
print "************************"
s_analysis_dir = analysis_dir.split("/")
pilot = s_analysis_dir[len(s_analysis_dir) - 2]
print s_analysis_dir
pass_thresh = 0
total = 0
ls_algs = os.listdir(analysis_dir)
for alg in ls_algs:
if (alg != "mrsfast"):
continue
print analysis_dir
alg_dir = "%s/%s" % (analysis_dir, alg)
print alg_dir
ls_indivs = os.listdir(alg_dir)
anal_by_indivs = {}
for indiv in ls_indivs:
ecs = []
corrs = []
ecs_corrs = {}
ecs_corrs["ecs"] = ecs
ecs_corrs["corrs"] = corrs
anal_by_indivs[indiv] = ecs_corrs
indiv_dir = "%s/%s" % (alg_dir, indiv)
ls_libs = os.listdir(indiv_dir)
for lib in ls_libs:
#if(lib[0:1] == "_"): continue
if (lib[0:1] == "_" or lib[0:3] == "do_" or lib[0] == "."):
continue
total += 1
lib_dir = "%s/%s" % (indiv_dir, lib)
if (get_lib_cov_corr(lib_dir) != None):
if (kgf.lib_pass_qc(lib_dir)):
pass_thresh += 1
(ec, corr) = get_lib_cov_corr(lib_dir)
ecs.append(ec)
corrs.append(corr)
do_plot = False
if do_plot:
figure(1)
indivs = []
for indiv in anal_by_indivs:
indivs.append(indiv)
ecs_corrs = anal_by_indivs[indiv]
ecs = np.array(ecs_corrs["ecs"])
corrs = np.array(ecs_corrs["corrs"])
marker = '.'
if (indiv[2:4] == "19" and pilot == "Pilot2"):
marker = '^'
if do_plot:
plot(ecs, corrs, marker)
thresh = kgf.get_lib_pass_thresh()
plot(np.array((0, 12)), np.array((thresh, thresh)), "r--")
indivs.append("threshold")
if (pilot == "Pilot2"):
legend(indivs, loc=4)
xlabel("effective coverage bp")
ylabel("correlation with bacs (X inc)")
title("%s - correlation vs ec - %d/%d meet threshold" %
(pilot, pass_thresh, total))
figurename = "%s/%s" % (analysis_dir,
"/_all_alg_analysis/ec_vs_corr_dim0.png")
if do_plot:
savefig(figurename, format='png')
print figurename
close(1)
def analyze_coverage_to_correlation(analysis_dir):
#anal_by_alg = {}
print "************************"
print "analyze_coverage_to_correlation"
print "************************"
s_analysis_dir = analysis_dir.split("/")
pilot = s_analysis_dir[len(s_analysis_dir)-2]
print s_analysis_dir
pass_thresh = 0;
total = 0
ls_algs = os.listdir(analysis_dir)
for alg in ls_algs:
if(alg != "mrsfast"):
continue
print analysis_dir
alg_dir = "%s/%s"%(analysis_dir,alg)
print alg_dir
ls_indivs = os.listdir(alg_dir)
anal_by_indivs = {}
for indiv in ls_indivs:
ecs = []
corrs = []
ecs_corrs = {}
ecs_corrs["ecs"] = ecs
ecs_corrs["corrs"] = corrs
anal_by_indivs[indiv] = ecs_corrs
indiv_dir = "%s/%s"%(alg_dir,indiv)
ls_libs = os.listdir(indiv_dir)
for lib in ls_libs:
#if(lib[0:1] == "_"): continue
if(lib[0:1] == "_" or lib[0:3]=="do_" or lib[0]=="."):continue
total+=1
lib_dir = "%s/%s"%(indiv_dir,lib)
if(get_lib_cov_corr(lib_dir) != None):
if(kgf.lib_pass_qc(lib_dir)):
pass_thresh+=1
(ec,corr) = get_lib_cov_corr(lib_dir)
ecs.append(ec)
corrs.append(corr)
do_plot = False
if do_plot:
figure(1)
indivs = []
for indiv in anal_by_indivs:
indivs.append(indiv)
ecs_corrs = anal_by_indivs[indiv]
ecs = np.array(ecs_corrs["ecs"])
corrs = np.array(ecs_corrs["corrs"])
marker = '.'
if(indiv[2:4] == "19" and pilot=="Pilot2"):
marker = '^'
if do_plot:
plot(ecs,corrs,marker)
thresh = kgf.get_lib_pass_thresh()
plot(np.array((0,12)),np.array((thresh,thresh)),"r--")
indivs.append("threshold")
if(pilot=="Pilot2"):
legend(indivs,loc=4)
xlabel("effective coverage bp")
ylabel("correlation with bacs (X inc)")
title("%s - correlation vs ec - %d/%d meet threshold"%(pilot,pass_thresh,total))
figurename = "%s/%s"%(analysis_dir,"/_all_alg_analysis/ec_vs_corr_dim0.png")
if do_plot:
savefig(figurename,format='png')
print figurename
close(1)