Skip to content

Commit

Permalink
Merge branch 'hotfix/3.5.5'
Browse files Browse the repository at this point in the history
  • Loading branch information
@blex-max
blex-max committed Feb 1, 2024
2 parents f2a469a + 9026421 commit fc6f91e
Show file tree
Hide file tree
Showing 3 changed files with 175 additions and 168 deletions.
3 changes: 2 additions & 1 deletion .gitignore
View file
Original file line number Diff line number Diff line change
@@ -1 +1,2 @@
bugfix-scratch/
bugfix-scratch/
bug-data/
328 changes: 167 additions & 161 deletions R/nanoseq_results_plotter.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -148,6 +148,11 @@ n_variants <- burdens[ismasked == 0][isvariant == 1]$count
n_reference <- burdens[ismasked == 0][isvariant == 0]$count
n_unique <- nrow(unique_variants[ismasked == 0])

if(length(n_reference) == 0) {
message("0 reference calls. Exiting…\n")
quit(save = "no", status = 1)
}

if (length(n_variants) == 0) { n_variants = 0; }
metrics <- data.frame("Metric" = c("total variants", "unique variants",
"total variant + reference", "uncorrected burden", "physical coverage"),
Expand Down Expand Up @@ -568,181 +573,182 @@ write.table(tosave,file=paste(out_name, ".obs_burdens.pre_vs_post_masking.tsv",
# Mismatches
# To understand error profile - damage during library preparation
mismatches = read.table(paste(dirname, 'mismatches.csv', sep = "/"), sep = ",", header = T)
mismatches$sub = paste(unlist(sapply(mismatches$mismatch, function(x) unlist(strsplit(x, ""))[2])), ">", unlist(sapply(mismatches$mismatch, function(x) unlist(strsplit(x, ""))[5])), sep = "")
mismatches = mismatches[which(mismatches$ismasked == 0),]
mismatches = mismatches[which(mismatches$mismatch != "UNKNOWN"),]
total = nrow(mismatches)

pdf(width = 5, height = 5, file = paste(out_name, ".mismatches.subst_asym.pdf", sep = ""))

subs = c("C>A", "G>T", "C>G", "G>C", "C>T", "G>A", "T>A", "A>T", "T>C", "A>G", "T>G", "A>C")
colors = rep(c(rgb(34 / 255, 159 / 255, 198 / 255), rgb(26 / 255, 26 / 255, 26 / 255),
rgb(201 / 255, 93 / 255, 94 / 255), rgb(178 / 255, 182 / 255, 180 / 255),
rgb(153 / 255, 208 / 255, 62 / 255), rgb(217 / 255, 190 / 255, 217 / 255)), each = 2)
par(mar = c(5, 5, 5, 5))
counts = table(mismatches$sub)
tmp_ = names(counts)
counts = as.vector(counts)
names(counts) = tmp_
counts[setdiff(subs, names(counts))] = 0
counts = counts[intersect(subs, names(counts))]
counts = counts[subs]
write.table(counts, file = paste(out_name, ".mismatches.subst_asym.tsv", sep = ""), quote = F, sep = "\t", row.names = F, col.names = F)

tick_10 = 10 * total / 100
bar = barplot(counts, las = 2, col = colors,
xlab = "Substitution (read strand)", ylab = "Number of mutations", border = "NA",
ylim = c(0, max(counts) + 0.2 * max(counts)))
axis(side = 4, at = c(0, tick_10, tick_10 * 2, tick_10 * 3), labels = c("0", "10", "20", "30"))
mtext("% of mutations", side = 4, line = 2)
# Add significance indicators
pvalues = vector()
for (i in c(1, 3, 5, 7, 9, 11)) {
j = i + 1
if (counts[i] + counts[j] > 0) {
pvalues[subs[i]] = binom.test(counts[i], counts[i] + counts[j])$p.value
} else {
pvalues[subs[i]] = 1
}
}
write.table(pvalues, file = paste(out_name, ".mismatches.subst_asym.pvals", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
qvalues = p.adjust(pvalues, method = "BH")
for (i in c(1, 3, 5, 7, 9, 11)) {
indicator = ""
if (qvalues[subs[i]] < 0.01) {
indicator = "**"
} else if (qvalues[subs[i]] < 0.1) {
indicator = "*"
if(nrow(mismatches) > 0) {
mismatches$sub = paste(unlist(sapply(mismatches$mismatch, function(x) unlist(strsplit(x, ""))[2])), ">", unlist(sapply(mismatches$mismatch, function(x) unlist(strsplit(x, ""))[5])), sep = "")
mismatches = mismatches[which(mismatches$ismasked == 0),]
mismatches = mismatches[which(mismatches$mismatch != "UNKNOWN"),]
total = nrow(mismatches)

pdf(width = 5, height = 5, file = paste(out_name, ".mismatches.subst_asym.pdf", sep = ""))

subs = c("C>A", "G>T", "C>G", "G>C", "C>T", "G>A", "T>A", "A>T", "T>C", "A>G", "T>G", "A>C")
colors = rep(c(rgb(34 / 255, 159 / 255, 198 / 255), rgb(26 / 255, 26 / 255, 26 / 255),
rgb(201 / 255, 93 / 255, 94 / 255), rgb(178 / 255, 182 / 255, 180 / 255),
rgb(153 / 255, 208 / 255, 62 / 255), rgb(217 / 255, 190 / 255, 217 / 255)), each = 2)
par(mar = c(5, 5, 5, 5))
counts = table(mismatches$sub)
tmp_ = names(counts)
counts = as.vector(counts)
names(counts) = tmp_
counts[setdiff(subs, names(counts))] = 0
counts = counts[intersect(subs, names(counts))]
counts = counts[subs]
write.table(counts, file = paste(out_name, ".mismatches.subst_asym.tsv", sep = ""), quote = F, sep = "\t", row.names = F, col.names = F)

tick_10 = 10 * total / 100
bar = barplot(counts, las = 2, col = colors,
xlab = "Substitution (read strand)", ylab = "Number of mutations", border = "NA",
ylim = c(0, max(counts) + 0.2 * max(counts)))
axis(side = 4, at = c(0, tick_10, tick_10 * 2, tick_10 * 3), labels = c("0", "10", "20", "30"))
mtext("% of mutations", side = 4, line = 2)
# Add significance indicators
pvalues = vector()
for (i in c(1, 3, 5, 7, 9, 11)) {
j = i + 1
if (counts[i] + counts[j] > 0) {
pvalues[subs[i]] = binom.test(counts[i], counts[i] + counts[j])$p.value
} else {
pvalues[subs[i]] = 1
}
}
text((bar[i] + bar[i + 1]) / 2, max(counts[i], counts[i + 1]), indicator, pos = 3)
}
dev.off()

#########
# Now the trinuc profiles
pdf(width = 15, height = 8, file = paste(out_name, ".mismatches.trinuc-profile.pdf", sep = ""))
par(mfrow = c(2, 1))
par(mar = c(4, 6, 2, 2))

complement_subs2pyr = function(sub) {
first = unlist(strsplit(sub, ""))[1]
from = unlist(strsplit(sub, ""))[2]
third = unlist(strsplit(sub, ""))[3]
to = unlist(strsplit(sub, ""))[5]
complement = vector()
complement[c("A", "C", "G", "T")] = c("T", "G", "C", "A")
pyr = c("C", "T")
if (from %in% pyr) {
return(sub)
} else {
sub = paste(complement[third], complement[from], complement[first], ">", complement[to], sep = "")
return(sub)
write.table(pvalues, file = paste(out_name, ".mismatches.subst_asym.pvals", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
qvalues = p.adjust(pvalues, method = "BH")
for (i in c(1, 3, 5, 7, 9, 11)) {
indicator = ""
if (qvalues[subs[i]] < 0.01) {
indicator = "**"
} else if (qvalues[subs[i]] < 0.1) {
indicator = "*"
}
text((bar[i] + bar[i + 1]) / 2, max(counts[i], counts[i + 1]), indicator, pos = 3)
}
}
dev.off()

mismatches$mismatch_pyr = sapply(mismatches$mismatch, function(x) complement_subs2pyr(x))
#########
# Now the trinuc profiles
pdf(width = 15, height = 8, file = paste(out_name, ".mismatches.trinuc-profile.pdf", sep = ""))
par(mfrow = c(2, 1))
par(mar = c(4, 6, 2, 2))

colours = rep(c("deepskyblue", "black", "firebrick2", "gray", "darkolivegreen3", "rosybrown2"), each = 16)
sub_vec = c("C>A", "C>G", "C>T", "T>A", "T>C", "T>G")
ctx_vec = paste(rep(c("A", "C", "G", "T"), each = 4), rep(c("A", "C", "G", "T"), times = 4), sep = "-")
full_vec = paste(rep(sub_vec, each = 16), rep(ctx_vec, times = 6), sep = ",")
xstr = paste(substr(full_vec, 5, 5), substr(full_vec, 1, 1), substr(full_vec, 7, 7), sep = "")
ordered_names = paste(xstr, ">", rep(c("A", "G", "T", "A", "C", "G"), each = 16), sep = "")
tmp_ = table(mismatches[which(mismatches$mismatch == mismatches$mismatch_pyr), "mismatch"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
complement_subs2pyr = function(sub) {
first = unlist(strsplit(sub, ""))[1]
from = unlist(strsplit(sub, ""))[2]
third = unlist(strsplit(sub, ""))[3]
to = unlist(strsplit(sub, ""))[5]
complement = vector()
complement[c("A", "C", "G", "T")] = c("T", "G", "C", "A")
pyr = c("C", "T")
if (from %in% pyr) {
return(sub)
} else {
sub = paste(complement[third], complement[from], complement[first], ">", complement[to], sep = "")
return(sub)
}
}

# Observed counts
y = tri_obs;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Observed Pyrimidine mismatches counts", main = "Observed Pyirimidine mismatches counts")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Pyrimidine.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
tri_obs_pyr = tri_obs
mismatches$mismatch_pyr = sapply(mismatches$mismatch, function(x) complement_subs2pyr(x))

tmp_ = table(mismatches[which(mismatches$mismatch != mismatches$mismatch_pyr), "mismatch_pyr"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
colours = rep(c("deepskyblue", "black", "firebrick2", "gray", "darkolivegreen3", "rosybrown2"), each = 16)
sub_vec = c("C>A", "C>G", "C>T", "T>A", "T>C", "T>G")
ctx_vec = paste(rep(c("A", "C", "G", "T"), each = 4), rep(c("A", "C", "G", "T"), times = 4), sep = "-")
full_vec = paste(rep(sub_vec, each = 16), rep(ctx_vec, times = 6), sep = ",")
xstr = paste(substr(full_vec, 5, 5), substr(full_vec, 1, 1), substr(full_vec, 7, 7), sep = "")
ordered_names = paste(xstr, ">", rep(c("A", "G", "T", "A", "C", "G"), each = 16), sep = "")
tmp_ = table(mismatches[which(mismatches$mismatch == mismatches$mismatch_pyr), "mismatch"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]

# Observed counts
y = tri_obs;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Observed Purine mismatches counts", main = "Observed Purine mismatches counts")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Purine.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
tri_obs = tri_obs_pyr + tri_obs
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Both.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
# Observed counts
y = tri_obs;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Observed Pyrimidine mismatches counts", main = "Observed Pyirimidine mismatches counts")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Pyrimidine.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
tri_obs_pyr = tri_obs

tmp_ = table(mismatches[which(mismatches$mismatch != mismatches$mismatch_pyr), "mismatch_pyr"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]

##########################################################################################
# Now error rates based on SSC, by each of 96 trinuc-sub channels and global
tmp_ = table(mismatches[which(mismatches$mismatch == mismatches$mismatch_pyr), "mismatch"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
tri_obs_pyr = tri_obs
tmp_ = table(mismatches[which(mismatches$mismatch != mismatches$mismatch_pyr), "mismatch_pyr"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
tri_obs_pur = tri_obs
# Observed counts
y = tri_obs;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Observed Purine mismatches counts", main = "Observed Purine mismatches counts")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Purine.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
tri_obs = tri_obs_pyr + tri_obs
write.table(tri_obs, file = paste(out_name, ".SSC-mismatches-Both.triprofiles.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)

pdf(width = 15, height = 4, file = paste(out_name, ".DSC_estimated_error_rates.pdf", sep = ""))
par(mar = c(4, 6, 2, 2))
tris = sapply(names(tri_obs), function(x) unlist(strsplit(x, ">"))[1])
y = (tri_obs_pyr / (tri_bg[tris] / 2)) * (tri_obs_pur / (tri_bg[tris] / 2));
maxy = max(y)
##y = (tri_obs_pyr/(tri_bg[tris])) * (tri_obs_pur/(tri_bg[tris])); maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Error rates", main = "Error rates")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()

##########################################################################################
# Now error rates based on SSC, by each of 96 trinuc-sub channels and global
tmp_ = table(mismatches[which(mismatches$mismatch == mismatches$mismatch_pyr), "mismatch"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
tri_obs_pyr = tri_obs
tmp_ = table(mismatches[which(mismatches$mismatch != mismatches$mismatch_pyr), "mismatch_pyr"])
tri_obs = as.vector(tmp_)
names(tri_obs) = names(tmp_)
tri_obs[setdiff(ordered_names, names(tri_obs))] = 0
tri_obs = tri_obs[ordered_names]
tri_obs_pur = tri_obs

# Global:
# Number of errors per channel:
errors_per_channel = y * tri_bg[tris]
total_error_rate = sum(errors_per_channel) / sum(tri_bg)
pdf(width = 15, height = 4, file = paste(out_name, ".DSC_estimated_error_rates.pdf", sep = ""))
par(mar = c(4, 6, 2, 2))
tris = sapply(names(tri_obs), function(x) unlist(strsplit(x, ">"))[1])
y = (tri_obs_pyr / (tri_bg[tris] / 2)) * (tri_obs_pur / (tri_bg[tris] / 2));
maxy = max(y)
##y = (tri_obs_pyr/(tri_bg[tris])) * (tri_obs_pur/(tri_bg[tris])); maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Error rates", main = "Error rates")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()

pdf(width = 15, height = 4, file = paste(out_name, ".DSC_errors_per_channel.pdf", sep = ""))
par(mar = c(4, 6, 2, 2))
tris = sapply(names(tri_obs), function(x) unlist(strsplit(x, ">"))[1])
y = errors_per_channel;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Predicted errors", main = "Predicted errors")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()

output = list()
output[["errors_per_channel"]] = errors_per_channel
output[["error_rates_per_channel"]] = y
output[["total_error_rate"]] = total_error_rate
output[["total_errors"]] = sum(errors_per_channel)
kkk = vector()
kkk["total_error_rate"] = total_error_rate
kkk["total_errors"] = output[["total_errors"]]
write.table(kkk, file = paste(out_name, ".estimated_error_rates.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
# Global:
# Number of errors per channel:
errors_per_channel = y * tri_bg[tris]
total_error_rate = sum(errors_per_channel) / sum(tri_bg)

pdf(width = 15, height = 4, file = paste(out_name, ".DSC_errors_per_channel.pdf", sep = ""))
par(mar = c(4, 6, 2, 2))
tris = sapply(names(tri_obs), function(x) unlist(strsplit(x, ">"))[1])
y = errors_per_channel;
maxy = max(y)
h = barplot(y, las = 2, col = colours, border = NA, ylim = c(0, maxy * 1.5), space = 1, cex.names = 0.6, names.arg = xstr, ylab = "Predicted errors", main = "Predicted errors")
for (j in 1:length(sub_vec)) {
xpos = h[c((j - 1) * 16 + 1, j * 16)]
rect(xpos[1] - 0.5, maxy * 1.2, xpos[2] + 0.5, maxy * 1.3, border = NA, col = colours[j * 16])
text(x = mean(xpos), y = maxy * 1.3, pos = 3, label = sub_vec[j])
}
dev.off()

output = list()
output[["errors_per_channel"]] = errors_per_channel
output[["error_rates_per_channel"]] = y
output[["total_error_rate"]] = total_error_rate
output[["total_errors"]] = sum(errors_per_channel)
kkk = vector()
kkk["total_error_rate"] = total_error_rate
kkk["total_errors"] = output[["total_errors"]]
write.table(kkk, file = paste(out_name, ".estimated_error_rates.tsv", sep = ""), quote = F, sep = "\t", row.names = T, col.names = F)
}
12 changes: 6 additions & 6 deletions python/runNanoSeq.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -423,9 +423,9 @@ def runCommand(command):
print("\nExecuting: %s\n" % ijob)
p = subprocess.Popen(
ijob, shell=True, stdout=subprocess.DEVNULL, stderr=subprocess.PIPE)
p.wait()
std: tuple = p.communicate()
if (p.returncode != 0):
error = p.stderr.read().decode()
error = std[1].decode()
sys.stderr.write("\n!Error processing: %s\n" % ijob)
raise ValueError(error)
return
Expand Down Expand Up @@ -1205,10 +1205,10 @@ def vcfHeader(args):
ifile = tmpDir+"/indel/%s.indel.filtered.vcf.gz" % (i+1)
if ( os.stat(ifile).st_size == 0 ) : continue
vcf2Merge.append(ifile)
cmd = "cp %s merged.vcf.gz ;"% vcf2Merge.pop(0)
for ifile in vcf2Merge[1:]:
cmd += " bcftools concat --no-version -Oz -o tmp.xxx.vcf.gz merged.vcf.gz %s; mv tmp.xxx.vcf.gz merged.vcf.gz; "%ifile
cmd += "bcftools sort -Oz -o %s/post/%s.indel.vcf.gz merged.vcf.gz;" % (tmpDir, args.name)
cmd = "cp %s %s/indel/merged.vcf.gz ;" % (vcf2Merge.pop(0), tmpDir)
for ifile in vcf2Merge[0:]:
cmd += "bcftools concat --no-version -Oz -o %s/indel/tmp.xxx.vcf.gz %s/indel/merged.vcf.gz %s; mv %s/indel/tmp.xxx.vcf.gz %s/indel/merged.vcf.gz; " % (tmpDir, tmpDir, ifile, tmpDir, tmpDir)
cmd += "bcftools sort -Oz -o %s/post/%s.indel.vcf.gz %s/indel/merged.vcf.gz;" % (tmpDir, args.name, tmpDir)
cmd += "bcftools index -t -f %s/post/%s.indel.vcf.gz " % (
tmpDir, args.name)
runCommand(cmd)
Expand Down

0 comments on commit fc6f91e

Please sign in to comment.