2025_exp_4_alginate_DEG
Nour El Husseini
2025
1 Overview
| batch | condition | media | treatment | induced | exp_4 | |
|---|---|---|---|---|---|---|
| SK064 | 1 | LB-IPTG | LB | DMSO | yes | control |
| SK065 | 2 | LB-IPTG | LB | DMSO | yes | control |
| SK066 | 3 | LB-IPTG | LB | DMSO | yes | control |
| SK070 | 1 | LB-EbO-IPTG | LB | EbO | yes | treatment |
| SK071 | 2 | LB-EbO-IPTG | LB | EbO | yes | treatment |
| SK072 | 3 | LB-EbO-IPTG | LB | EbO | yes | treatment |
2 Project variables
## batch condition media treatment induced exp_4
## SK064 1 LB-IPTG LB DMSO yes control
## SK065 2 LB-IPTG LB DMSO yes control
## SK066 3 LB-IPTG LB DMSO yes control
## SK070 1 LB-EbO-IPTG LB EbO yes treatment
## SK071 2 LB-EbO-IPTG LB EbO yes treatment
## SK072 3 LB-EbO-IPTG LB EbO yes treatment## [1] "The control is: LB-IPTG"## [1] "The treatment is: LB-EbO-IPTG"# Load raw count data
rawcounts <- read.csv("../../raw-data/htseq_SK_unfiltered_reverse_strand_count.csv",
row.names = 1, header = TRUE)
# Append a prefix to the batch and run columns to turn to character class
metadata[["batch"]] <- paste0("b", metadata[["batch"]])
metadata[["run"]] <- paste0("r", metadata[["run"]])
# Subset the rawcounts dataframe to only include samples that are present in experiment
samples <- rownames(metadata)
rawcounts <- select(rawcounts, all_of(samples))
# Ensure that the order of samples in metadata matches the order of columns in rawcounts
all(rownames(metadata) == colnames(rawcounts)) # Should return TRUE if order is correct## [1] TRUE# Define the design formula for DESeq2, including main condition and batch effect
design_formula <- ~ exp_4 + batch
# Create DESeq2 dataset object from the filtered count matrix and metadata
dds <- DESeqDataSetFromMatrix(
countData = rawcounts,
colData = metadata,
design = design_formula
)## Warning in DESeqDataSet(se, design = design, ignoreRank): some variables in
## design formula are characters, converting to factors3 Library size per sample
# Sum up the counts for each sample
rawcounts_sums <- colSums(rawcounts)
# Turn it into a data frame
data <- data.frame(
Category = names(rawcounts_sums),
Count = as.numeric(rawcounts_sums)
)
# Make sure sample order is consistent
data <- data %>%
mutate(Category = factor(Category, levels = sort(unique(Category))))
# Plot bar chart of library sizes
p <- ggplot(data, aes(x = Category, y = Count, fill = Category)) +
geom_bar(stat = "identity") +
# coord_flip() # flip if too many samples to read labels easily
labs(y = "Library size (M)", x = NULL) + # remove x label, rename y axis
# scale_fill_brewer(palette = "Dark2") + # optional color palette
scale_y_continuous(
labels = scales::label_number(scale = 1e-6), # show counts in millions
limits = c(0, (max(data$Count) + 1e6)) # add a bit of space on top
) +
theme(
panel.grid = element_blank(),
panel.background = element_blank(),
axis.text = element_text(size = text.size, angle = 45, hjust = 1), # slant labels
axis.title = element_text(size = axis.text.size),
legend.text = element_text(size = text.size),
legend.title = element_text(size = axis.text.size),
panel.border = element_rect(fill = NA),
legend.position = "none", # no legend needed since each bar is a different sample
axis.ticks = element_line(linewidth = 0.2),
legend.key = element_blank()
)
# Show the plot
p
# Save the plot as a PDF with today’s date and project name
ggsave(paste0(Sys.Date(), "_library_size_", exp, "_", treatment,".vs.", control, ".pdf"), device = "pdf")## Saving 7 x 5 in image4 Expression distribution
# Stack rawcounts into single vector
dat <- stack(rawcounts)
# Format expression distribution plot with the expression level and density
expression_dist <- ggplot(dat, aes(x = log2(values+1), fill=ind, label = ind )) +
geom_density(alpha = 0.5, position = "identity") +
ylab("Density") +
geom_textdensity(hjust = "ymax", vjust = 0,
text_only = TRUE, text_smoothing = 20) +
theme(panel.grid = element_blank(),
panel.background = element_blank(),
axis.text = element_text(size = text.size),
axis.title = element_text(size = axis.text.size),
legend.text = element_text(size = text.size),
legend.title = element_text(size = axis.text.size),
panel.border = element_rect(fill = NA),
legend.position = "none",
axis.ticks = element_line(linewidth = 0.2),
legend.key = element_blank()) +
coord_cartesian(xlim = c(0, 15))
expression_dist
#save figure
cairo_pdf(paste0(Sys.Date(), "_expression_dist_", exp, "_", treatment, ".vs.", control, ".pdf"))
expression_dist
dev.off() ## quartz_off_screen
## 25 Non-zero genes observed
# Prepare data by getting library sizes and non-zero counts
x <- colSums(rawcounts)
y <- colSums(rawcounts != 0)
df <- data.frame(x,y)
df$ind <- rownames(df)
# Calculate axis limits
x_min <- min(df$x)
x_max <- max(df$x)
y_min <- min(df$y)
y_max <- max(df$y)
# Non-zero plot
nonzero <- ggplot(data = df, aes(x = x, y = y, color = ind, label = ind)) +
geom_point(size = 3) +
geom_text(hjust = -0.1, vjust = -0.1, color = "black", size = 5) +
scale_y_continuous(limits=c(y_min, y_max)) +
scale_x_continuous(
limits = c(x_min-1e3, x_max+9e5),
labels = scales::comma_format(scale = 1e-6, accuracy = 1) # Format labels without decimals
) +
xlab("Millions of reads") + ylab("Number of non-zero genes observed") +
labs(color = "Sample") +
theme(panel.grid = element_blank(),
panel.background = element_blank(),
axis.text = element_text(size = text.size),
axis.title = element_text(size = axis.text.size),
legend.text = element_text(size = text.size),
legend.title = element_text(size = axis.text.size),
panel.border = element_rect(fill = NA),
legend.position = "none",
axis.ticks = element_line(linewidth = 0.2),
legend.key = element_blank()) +
coord_cartesian()
# Show nonzero plot
nonzero
# Save plot
cairo_pdf(paste0(Sys.Date(), "_non-zero_", exp, "_", treatment, ".vs.", control, ".pdf"))
nonzero
dev.off() ## quartz_off_screen
## 26 RPKM and CPM transformations
# Load GFF annotation
gff_annot <- load_gff_annotations("../../raw-data/paeruginosa_pa14.gff",
id_col = "gene_id", type = "gene")## Returning a df with 16 columns and 5979 rows.## seqnames start end width strand source type score phase
## gene1650835 chromosome 483 2027 1545 + PseudoCAP gene NA 0
## gene1650837 chromosome 2056 3159 1104 + PseudoCAP gene NA 0
## gene1650839 chromosome 3169 4278 1110 + PseudoCAP gene NA 0
## gene1650841 chromosome 4275 6695 2421 + PseudoCAP gene NA 0
## gene1650843 chromosome 7018 7791 774 - PseudoCAP gene NA 0
## gene1650845 chromosome 7803 8339 537 - PseudoCAP gene NA 0
## gene_id Name Dbxref Alias name Parent locus
## gene1650835 gene1650835 GeneID:4384099 PA14_00010 dnaA <NA>
## gene1650837 gene1650837 GeneID:4384100 PA14_00020 dnaN <NA>
## gene1650839 gene1650839 GeneID:4384101 PA14_00030 recF <NA>
## gene1650841 gene1650841 GeneID:4384102 PA14_00050 gyrB <NA>
## gene1650843 gene1650843 GeneID:4385186 PA14_00060 PA14_00060 <NA>
## gene1650845 gene1650845 GeneID:4385187 PA14_00070 PA14_00070 <NA>## SK064 SK065 SK066 SK070 SK071 SK072
## gene1650835 2234 2528 1719 857 1136 856
## gene1650837 1266 1754 1368 678 757 719
## gene1650839 722 770 657 272 312 208
## gene1650841 3671 4307 3347 2806 2782 2278
## gene1650843 436 435 313 144 154 121
## gene1650845 257 299 177 76 97 91## [1] 5979 16## [1] 5979 6## Mode FALSE TRUE
## logical 932 5047## Mode TRUE
## logical 5979# Create an expression object with counts, metadata, and gene info
expt <- create_expt(count_dataframe = rawcounts,
metadata = metadata,
gene_info = gff_annot)## Reading the sample metadata.## Did not find the column: sampleid.## The rownames do not appear numeric, using them.## The sample definitions comprises: 6 rows(samples) and 8 columns(metadata fields).## Matched 5979 annotations and counts.## Bringing together the count matrix and gene information.## Some annotations were lost in merging, setting them to 'undefined'.## Saving the expressionset to 'expt.rda'.## The final expressionset has 5979 features and 6 samples.# Normalize counts log2 CPM
cpm_normalized_exp <- normalize_expt(expt,
transform = "log2",
norm = "raw",
convert = "cpm",
row_min = "10")## transform_counts: Found 288 values equal to 0, adding 1 to the matrix.####### Calculate RPKM ########
# Grab gene lengths from GFF (should be in base pairs)
gene_lengths <- gff_annot$width
names(gene_lengths) <- rownames(gff_annot) # set gene names
# Make sure gene_lengths are in the same order as rawcounts rows
gene_lengths <- gene_lengths[rownames(rawcounts)]
# Make DGEList object for edgeR
dge <- DGEList(counts = rawcounts)
# Calculate RPKM (reads per kilobase per million)
rpkm <- rpkm(dge, gene.length = gene_lengths)
# Wrap RPKM in an experiment object too
expt_rpkm <- create_expt(count_dataframe = rpkm,
metadata = metadata,
gene_info = gff_annot)## Reading the sample metadata.## Did not find the column: sampleid.## The rownames do not appear numeric, using them.## The sample definitions comprises: 6 rows(samples) and 8 columns(metadata fields).## Matched 5979 annotations and counts.## Bringing together the count matrix and gene information.## Some annotations were lost in merging, setting them to 'undefined'.## Saving the expressionset to 'expt.rda'.## The final expressionset has 5979 features and 6 samples.# log2 transform the RPKM (again, no normalization beyond log)
rpkm_normalized_exp <- normalize_expt(expt_rpkm,
transform = "log2",
norm = "raw")## transform_counts: Found 288 values equal to 0, adding 1 to the matrix.# Get the actual expression matrices from the experiment objects
rpkm <- exprs(rpkm_normalized_exp)
cpm <- exprs(cpm_normalized_exp)
# Merge CPM and RPKM matrices with gene annotation info
cpm_annotated <- merge(gff_annot, cpm, by = 0)
rpkm_annotated <- merge(gff_annot, rpkm, by = 0)6.1 Heatmap of alginate genes
This generates a small heatmap of the alginate genes across samples to visualize expression.
# Loop through CPM and RPKM to plot both heatmaps
for (type in c("cpm", "rpkm")) {
# Define gene list
alg_operon_genes <- c(
"algA", "algF", "algJ", "algI", "algL", "algX", "algG", "algE",
"algK", "alg44", "alg8", "algD", "mucC", "mucB", "algU", "algW",
"algP", "algQ", "algR", "algZ", "algC", "algB"
)
# Get the right data per type of transformation
data_annot <- get(paste0(type, "_annotated"))
# Fix rownames using gene name (make unique)
rownames(data_annot) <- make.names(data_annot$name, unique = TRUE)
# Drop undesired metadata columns
cols_to_exclude <- c("Row.names","seqnames","start","end","width",
"strand","source","type","score","phase",
"gene_id","Name","Dbxref","Alias","name",
"Parent","locus")
keep <- !names(data_annot) %in% cols_to_exclude
expr_data <- data_annot[, keep]
# Extract matrix for alg genes
alg_genes <- expr_data[alg_operon_genes, , drop = FALSE]
mat <- alg_genes - rowMeans(alg_genes)
# Pull sample annotations
anno <- as.data.frame(metadata[, c("condition", "induced")])
colnames(anno) <- c("Condition", "Induced")
# Color settings for annotation
# ann_colors <- list(
# Condition = c("LB-DMSO" = "#D95F02", "LB-IPTG" = "#66A61E"),
# Induced = c("no" = "lightgrey", "yes" = "gray27")
# )
# Scale data by row (Z-score per gene)
scaled_mat <- t(scale(t(mat)))
# Build heatmap
ht <- Heatmap(
scaled_mat,
name = "Z-score",
top_annotation = HeatmapAnnotation(df = anno#,
#col = ann_colors
),
cluster_rows = TRUE,
cluster_columns = TRUE,
row_names_side = "left",
column_names_side = "bottom",
column_names_rot = 45,
row_names_gp = gpar(fontsize = 10),
column_names_gp = gpar(fontsize = 10),
show_row_dend = TRUE,
show_column_dend = TRUE
)
# Draw and save
draw(ht)
cairo_pdf(paste0(Sys.Date(), "_alg_heatmap_", type, "_", exp, "_", treatment, ".vs.", control, ".pdf"))
draw(ht)
dev.off()
}
7 DESeq
# Create DESeqDataSet
dds <- DESeqDataSetFromMatrix(
countData = rawcounts,
colData = metadata,
design = design_formula
)## Warning in DESeqDataSet(se, design = design, ignoreRank): some variables in
## design formula are characters, converting to factors## [1] "Number of genes before filtering:"## [1] 5979# Filter: keep genes with counts >=10 in at least 2 samples
print("Filtering genes with counts >=10 in at least 2 samples")## [1] "Filtering genes with counts >=10 in at least 2 samples"keep <- rowSums(counts(dds) >= 10) >= 2
dds_filtered <- dds[keep, ]
# Check number of genes after filtering
print("Number of genes after filtering:")## [1] "Number of genes after filtering:"## [1] 5725## estimating size factors## estimating dispersions## gene-wise dispersion estimates## mean-dispersion relationship## final dispersion estimates## fitting model and testing## [1] "Intercept" "exp_4_treatment_vs_control"
## [3] "batch_b2_vs_b1" "batch_b3_vs_b1"
8 Save DE tables
res <- results(dds_final, contrast = c("exp_4", "treatment", "control"))
resdf <- as.data.frame(res)
resdf$gene_id <- rownames(resdf)
cpm <- as.data.frame(cpm)
rpkm <- as.data.frame(rpkm)
# append type of value to the dataframe
colnames(rpkm) <- paste0(colnames(rpkm), "_rpkm")
colnames(cpm) <- paste0(colnames(cpm), "_cpm")
rawcounts2 <- rawcounts
colnames(rawcounts2) <- paste0(colnames(rawcounts2), "_rawcounts")
# ensure column gene_id is there
rpkm$gene_id <- rownames(rpkm)
cpm$gene_id <- rownames(cpm)
rawcounts2$gene_id <- rownames(rawcounts2)
gff_annot <- gff_annot[rownames(resdf),]
dim(resdf)## [1] 5725 7# merge with siggenes
resdf <- full_join(resdf, gff_annot, by = "gene_id", copy = FALSE)
head(resdf)## baseMean log2FoldChange lfcSE stat pvalue padj
## 1 1433.5096 -0.57542360 0.2027957 -2.8374549 0.0045474775 0.015614397
## 2 1022.8812 -0.40904226 0.2094370 -1.9530566 0.0508128832 0.111344077
## 3 447.8995 -0.84892945 0.2368050 -3.5849300 0.0003371686 0.001787407
## 4 3108.6397 0.08073561 0.2049896 0.3938522 0.6936901383 0.790701174
## 5 242.0401 -0.88329828 0.2425974 -3.6410049 0.0002715760 0.001493973
## 6 149.8484 -0.84289423 0.2750258 -3.0647824 0.0021782841 0.008387440
## gene_id seqnames start end width strand source type score phase
## 1 gene1650835 chromosome 483 2027 1545 + PseudoCAP gene NA 0
## 2 gene1650837 chromosome 2056 3159 1104 + PseudoCAP gene NA 0
## 3 gene1650839 chromosome 3169 4278 1110 + PseudoCAP gene NA 0
## 4 gene1650841 chromosome 4275 6695 2421 + PseudoCAP gene NA 0
## 5 gene1650843 chromosome 7018 7791 774 - PseudoCAP gene NA 0
## 6 gene1650845 chromosome 7803 8339 537 - PseudoCAP gene NA 0
## Name Dbxref Alias name Parent locus
## 1 GeneID:4384099 PA14_00010 dnaA <NA>
## 2 GeneID:4384100 PA14_00020 dnaN <NA>
## 3 GeneID:4384101 PA14_00030 recF <NA>
## 4 GeneID:4384102 PA14_00050 gyrB <NA>
## 5 GeneID:4385186 PA14_00060 PA14_00060 <NA>
## 6 GeneID:4385187 PA14_00070 PA14_00070 <NA>## locusId accession GI scaffoldId start stop strand Alias
## 1 2194572 YP_788156.1 116053721 4582 483 2027 + PA14_00010
## 2 2194573 YP_788157.1 116053722 4582 2056 3159 + PA14_00020
## 3 2194574 YP_788158.1 116053723 4582 3169 4278 + PA14_00030
## 4 2194575 YP_788159.1 116053724 4582 4275 6695 + PA14_00050
## 5 2194576 YP_788160.1 116053725 4582 7791 7018 - PA14_00060
## 6 2194577 YP_788161.1 116053726 4582 8339 7803 - PA14_00070
## name desc COG
## 1 dnaA chromosomal replication initiator protein DnaA (NCBI) COG593
## 2 dnaN DNA polymerase III, beta chain (NCBI) COG592
## 3 recF DNA replication and repair protein RecF (NCBI) COG1195
## 4 gyrB DNA gyrase subunit B (NCBI) COG187
## 5 PA14_00060 putative acyltransferase (NCBI) COG204
## 6 PA14_00070 putative histidinol-phosphatase (NCBI) COG241
## COGFun
## 1 L
## 2 L
## 3 L
## 4 L
## 5 I
## 6 E
## COGDesc
## 1 ATPase involved in DNA replication initiation
## 2 DNA polymerase sliding clamp subunit (PCNA homolog)
## 3 Recombinational DNA repair ATPase (RecF pathway)
## 4 Type IIA topoisomerase (DNA gyrase/topo II, topoisomerase IV), B subunit
## 5 1-acyl-sn-glycerol-3-phosphate acyltransferase
## 6 Histidinol phosphatase and related phosphatases
## TIGRFam
## 1 TIGR00362 chromosomal replication initiator protein DnaA [dnaA]
## 2 TIGR00663 DNA polymerase III, beta subunit [dnaN]
## 3 TIGR00611 DNA replication and repair protein RecF [recF]
## 4 TIGR01059 DNA gyrase, B subunit [gyrB]
## 5
## 6 TIGR01656 histidinol-phosphate phosphatase domain,TIGR01662 HAD hydrolase, family IIIA
## TIGRRoles
## 1 DNA metabolism:DNA replication, recombination, and repair
## 2 DNA metabolism:DNA replication, recombination, and repair
## 3 DNA metabolism:DNA replication, recombination, and repair
## 4 DNA metabolism:DNA replication, recombination, and repair
## 5
## 6 Unknown function:Enzymes of unknown specificity
## GO
## 1 GO:0006270,GO:0006275,GO:0003688,GO:0017111,GO:0005524
## 2 GO:0006260,GO:0003677,GO:0003893,GO:0008408,GO:0016449,GO:0019984,GO:0003889,GO:0003894,GO:0015999,GO:0016450,GO:0003890,GO:0003895,GO:0016000,GO:0016451,GO:0003891,GO:0016448,GO:0016452
## 3 GO:0006281,GO:0005694,GO:0005524,GO:0017111,GO:0003697
## 4 GO:0006304,GO:0006265,GO:0005694,GO:0003918,GO:0005524
## 5 GO:0008152,GO:0003841
## 6 GO:0000105,GO:0004401
## EC ECDesc
## 1
## 2 2.7.7.7 DNA-directed DNA polymerase.
## 3
## 4 5.99.1.3 DNA topoisomerase (ATP-hydrolyzing).
## 5 2.3.1.51 1-acylglycerol-3-phosphate O-acyltransferase.
## 6 3.1.3.-col_to_remove <- intersect(colnames(pa_go),colnames(resdf))
col_to_remove <- setdiff(col_to_remove, "Alias") # returns all except the second vector
pa_go <- pa_go %>% select(-all_of(col_to_remove))
resdf <- full_join(resdf, rpkm, by = "gene_id", copy = FALSE)
resdf <- full_join(resdf, cpm, by = "gene_id", copy = FALSE)
resdf <- full_join(resdf, rawcounts2, by = "gene_id", copy = FALSE)
#resdf <- full_join(resdf, pa_go, by = "Alias", copy = FALSE)
resdf <- merge(resdf, pa_go, by = "Alias", copy = FALSE)
head(resdf)## Alias baseMean log2FoldChange lfcSE stat pvalue
## 1 PA14_00010 1433.5096 -0.57542360 0.2027957 -2.8374549 0.0045474775
## 2 PA14_00020 1022.8812 -0.40904226 0.2094370 -1.9530566 0.0508128832
## 3 PA14_00030 447.8995 -0.84892945 0.2368050 -3.5849300 0.0003371686
## 4 PA14_00050 3108.6397 0.08073561 0.2049896 0.3938522 0.6936901383
## 5 PA14_00060 242.0401 -0.88329828 0.2425974 -3.6410049 0.0002715760
## 6 PA14_00070 149.8484 -0.84289423 0.2750258 -3.0647824 0.0021782841
## padj gene_id seqnames start end width strand source type
## 1 0.015614397 gene1650835 chromosome 483 2027 1545 + PseudoCAP gene
## 2 0.111344077 gene1650837 chromosome 2056 3159 1104 + PseudoCAP gene
## 3 0.001787407 gene1650839 chromosome 3169 4278 1110 + PseudoCAP gene
## 4 0.790701174 gene1650841 chromosome 4275 6695 2421 + PseudoCAP gene
## 5 0.001493973 gene1650843 chromosome 7018 7791 774 - PseudoCAP gene
## 6 0.008387440 gene1650845 chromosome 7803 8339 537 - PseudoCAP gene
## score phase Name Dbxref name Parent locus SK064_rpkm SK065_rpkm
## 1 NA 0 GeneID:4384099 dnaA <NA> 8.176179 8.165287
## 2 NA 0 GeneID:4384100 dnaN <NA> 7.842995 8.122956
## 3 NA 0 GeneID:4384101 recF <NA> 7.029755 6.934080
## 4 NA 0 GeneID:4384102 gyrB <NA> 8.244496 8.285574
## 5 NA 0 GeneID:4385186 PA14_00060 <NA> 6.823946 6.633153
## 6 NA 0 GeneID:4385187 PA14_00070 <NA> 6.591050 6.619830
## SK066_rpkm SK070_rpkm SK071_rpkm SK072_rpkm SK064_cpm SK065_cpm SK066_cpm
## 1 7.893669 6.967659 7.171523 7.071258 8.802026 8.791120 8.519135
## 2 8.048415 7.113400 7.071516 7.302903 7.985143 8.265208 8.190642
## 3 6.988440 5.803510 5.800306 5.528118 7.179220 7.083470 7.137873
## 4 8.205787 8.024788 7.812078 7.830929 9.517304 9.558461 9.478520
## 5 6.444121 5.414295 5.312490 5.272893 6.458065 6.267796 6.079357
## 6 6.152874 5.030312 5.176708 5.386342 5.706889 5.735417 5.273245
## SK070_cpm SK071_cpm SK072_cpm SK064_rawcounts SK065_rawcounts SK066_rawcounts
## 1 7.591194 7.795595 7.695076 2234 2528 1719
## 2 7.255158 7.213245 7.444783 1266 1754 1368
## 3 5.951508 5.948298 5.675576 722 770 657
## 4 9.297136 9.083908 9.102808 3671 4307 3347
## 5 5.054545 4.953457 4.914153 436 435 313
## 6 4.170875 4.313689 4.518773 257 299 177
## SK070_rawcounts SK071_rawcounts SK072_rawcounts locusId accession GI
## 1 857 1136 856 2194572 YP_788156.1 116053721
## 2 678 757 719 2194573 YP_788157.1 116053722
## 3 272 312 208 2194574 YP_788158.1 116053723
## 4 2806 2782 2278 2194575 YP_788159.1 116053724
## 5 144 154 121 2194576 YP_788160.1 116053725
## 6 76 97 91 2194577 YP_788161.1 116053726
## scaffoldId stop desc COG
## 1 4582 2027 chromosomal replication initiator protein DnaA (NCBI) COG593
## 2 4582 3159 DNA polymerase III, beta chain (NCBI) COG592
## 3 4582 4278 DNA replication and repair protein RecF (NCBI) COG1195
## 4 4582 6695 DNA gyrase subunit B (NCBI) COG187
## 5 4582 7018 putative acyltransferase (NCBI) COG204
## 6 4582 7803 putative histidinol-phosphatase (NCBI) COG241
## COGFun
## 1 L
## 2 L
## 3 L
## 4 L
## 5 I
## 6 E
## COGDesc
## 1 ATPase involved in DNA replication initiation
## 2 DNA polymerase sliding clamp subunit (PCNA homolog)
## 3 Recombinational DNA repair ATPase (RecF pathway)
## 4 Type IIA topoisomerase (DNA gyrase/topo II, topoisomerase IV), B subunit
## 5 1-acyl-sn-glycerol-3-phosphate acyltransferase
## 6 Histidinol phosphatase and related phosphatases
## TIGRFam
## 1 TIGR00362 chromosomal replication initiator protein DnaA [dnaA]
## 2 TIGR00663 DNA polymerase III, beta subunit [dnaN]
## 3 TIGR00611 DNA replication and repair protein RecF [recF]
## 4 TIGR01059 DNA gyrase, B subunit [gyrB]
## 5
## 6 TIGR01656 histidinol-phosphate phosphatase domain,TIGR01662 HAD hydrolase, family IIIA
## TIGRRoles
## 1 DNA metabolism:DNA replication, recombination, and repair
## 2 DNA metabolism:DNA replication, recombination, and repair
## 3 DNA metabolism:DNA replication, recombination, and repair
## 4 DNA metabolism:DNA replication, recombination, and repair
## 5
## 6 Unknown function:Enzymes of unknown specificity
## GO
## 1 GO:0006270,GO:0006275,GO:0003688,GO:0017111,GO:0005524
## 2 GO:0006260,GO:0003677,GO:0003893,GO:0008408,GO:0016449,GO:0019984,GO:0003889,GO:0003894,GO:0015999,GO:0016450,GO:0003890,GO:0003895,GO:0016000,GO:0016451,GO:0003891,GO:0016448,GO:0016452
## 3 GO:0006281,GO:0005694,GO:0005524,GO:0017111,GO:0003697
## 4 GO:0006304,GO:0006265,GO:0005694,GO:0003918,GO:0005524
## 5 GO:0008152,GO:0003841
## 6 GO:0000105,GO:0004401
## EC ECDesc
## 1
## 2 2.7.7.7 DNA-directed DNA polymerase.
## 3
## 4 5.99.1.3 DNA topoisomerase (ATP-hydrolyzing).
## 5 2.3.1.51 1-acylglycerol-3-phosphate O-acyltransferase.
## 6 3.1.3.-9 Save DEG list
## [1] 5718 54## Alias baseMean log2FoldChange lfcSE stat pvalue
## 1 PA14_00010 1433.5096 -0.57542360 0.2027957 -2.8374549 0.0045474775
## 2 PA14_00020 1022.8812 -0.40904226 0.2094370 -1.9530566 0.0508128832
## 3 PA14_00030 447.8995 -0.84892945 0.2368050 -3.5849300 0.0003371686
## 4 PA14_00050 3108.6397 0.08073561 0.2049896 0.3938522 0.6936901383
## 5 PA14_00060 242.0401 -0.88329828 0.2425974 -3.6410049 0.0002715760
## 6 PA14_00070 149.8484 -0.84289423 0.2750258 -3.0647824 0.0021782841
## padj gene_id seqnames start end width strand source type
## 1 0.015614397 gene1650835 chromosome 483 2027 1545 + PseudoCAP gene
## 2 0.111344077 gene1650837 chromosome 2056 3159 1104 + PseudoCAP gene
## 3 0.001787407 gene1650839 chromosome 3169 4278 1110 + PseudoCAP gene
## 4 0.790701174 gene1650841 chromosome 4275 6695 2421 + PseudoCAP gene
## 5 0.001493973 gene1650843 chromosome 7018 7791 774 - PseudoCAP gene
## 6 0.008387440 gene1650845 chromosome 7803 8339 537 - PseudoCAP gene
## score phase Name Dbxref name Parent locus SK064_rpkm SK065_rpkm
## 1 NA 0 GeneID:4384099 dnaA <NA> 8.176179 8.165287
## 2 NA 0 GeneID:4384100 dnaN <NA> 7.842995 8.122956
## 3 NA 0 GeneID:4384101 recF <NA> 7.029755 6.934080
## 4 NA 0 GeneID:4384102 gyrB <NA> 8.244496 8.285574
## 5 NA 0 GeneID:4385186 PA14_00060 <NA> 6.823946 6.633153
## 6 NA 0 GeneID:4385187 PA14_00070 <NA> 6.591050 6.619830
## SK066_rpkm SK070_rpkm SK071_rpkm SK072_rpkm SK064_cpm SK065_cpm SK066_cpm
## 1 7.893669 6.967659 7.171523 7.071258 8.802026 8.791120 8.519135
## 2 8.048415 7.113400 7.071516 7.302903 7.985143 8.265208 8.190642
## 3 6.988440 5.803510 5.800306 5.528118 7.179220 7.083470 7.137873
## 4 8.205787 8.024788 7.812078 7.830929 9.517304 9.558461 9.478520
## 5 6.444121 5.414295 5.312490 5.272893 6.458065 6.267796 6.079357
## 6 6.152874 5.030312 5.176708 5.386342 5.706889 5.735417 5.273245
## SK070_cpm SK071_cpm SK072_cpm SK064_rawcounts SK065_rawcounts SK066_rawcounts
## 1 7.591194 7.795595 7.695076 2234 2528 1719
## 2 7.255158 7.213245 7.444783 1266 1754 1368
## 3 5.951508 5.948298 5.675576 722 770 657
## 4 9.297136 9.083908 9.102808 3671 4307 3347
## 5 5.054545 4.953457 4.914153 436 435 313
## 6 4.170875 4.313689 4.518773 257 299 177
## SK070_rawcounts SK071_rawcounts SK072_rawcounts locusId accession GI
## 1 857 1136 856 2194572 YP_788156.1 116053721
## 2 678 757 719 2194573 YP_788157.1 116053722
## 3 272 312 208 2194574 YP_788158.1 116053723
## 4 2806 2782 2278 2194575 YP_788159.1 116053724
## 5 144 154 121 2194576 YP_788160.1 116053725
## 6 76 97 91 2194577 YP_788161.1 116053726
## scaffoldId stop desc COG
## 1 4582 2027 chromosomal replication initiator protein DnaA (NCBI) COG593
## 2 4582 3159 DNA polymerase III, beta chain (NCBI) COG592
## 3 4582 4278 DNA replication and repair protein RecF (NCBI) COG1195
## 4 4582 6695 DNA gyrase subunit B (NCBI) COG187
## 5 4582 7018 putative acyltransferase (NCBI) COG204
## 6 4582 7803 putative histidinol-phosphatase (NCBI) COG241
## COGFun
## 1 L
## 2 L
## 3 L
## 4 L
## 5 I
## 6 E
## COGDesc
## 1 ATPase involved in DNA replication initiation
## 2 DNA polymerase sliding clamp subunit (PCNA homolog)
## 3 Recombinational DNA repair ATPase (RecF pathway)
## 4 Type IIA topoisomerase (DNA gyrase/topo II, topoisomerase IV), B subunit
## 5 1-acyl-sn-glycerol-3-phosphate acyltransferase
## 6 Histidinol phosphatase and related phosphatases
## TIGRFam
## 1 TIGR00362 chromosomal replication initiator protein DnaA [dnaA]
## 2 TIGR00663 DNA polymerase III, beta subunit [dnaN]
## 3 TIGR00611 DNA replication and repair protein RecF [recF]
## 4 TIGR01059 DNA gyrase, B subunit [gyrB]
## 5
## 6 TIGR01656 histidinol-phosphate phosphatase domain,TIGR01662 HAD hydrolase, family IIIA
## TIGRRoles
## 1 DNA metabolism:DNA replication, recombination, and repair
## 2 DNA metabolism:DNA replication, recombination, and repair
## 3 DNA metabolism:DNA replication, recombination, and repair
## 4 DNA metabolism:DNA replication, recombination, and repair
## 5
## 6 Unknown function:Enzymes of unknown specificity
## GO
## 1 GO:0006270,GO:0006275,GO:0003688,GO:0017111,GO:0005524
## 2 GO:0006260,GO:0003677,GO:0003893,GO:0008408,GO:0016449,GO:0019984,GO:0003889,GO:0003894,GO:0015999,GO:0016450,GO:0003890,GO:0003895,GO:0016000,GO:0016451,GO:0003891,GO:0016448,GO:0016452
## 3 GO:0006281,GO:0005694,GO:0005524,GO:0017111,GO:0003697
## 4 GO:0006304,GO:0006265,GO:0005694,GO:0003918,GO:0005524
## 5 GO:0008152,GO:0003841
## 6 GO:0000105,GO:0004401
## EC ECDesc
## 1
## 2 2.7.7.7 DNA-directed DNA polymerase.
## 3
## 4 5.99.1.3 DNA topoisomerase (ATP-hydrolyzing).
## 5 2.3.1.51 1-acylglycerol-3-phosphate O-acyltransferase.
## 6 3.1.3.-## Mode FALSE
## logical 5718## Mode FALSE TRUE
## logical 5607 111sig_genes <- resdf[abs(resdf$log2FoldChange) >= logfc_cutoff & resdf$padj < pval_cutoff, ] %>% drop_na(log2FoldChange)
sig_genes$direction <- ifelse(sig_genes$log2FoldChange > 0, "up", "down")
summary(is.na(sig_genes$log2FoldChange))## Mode FALSE
## logical 88## Mode FALSE
## logical 88## [1] 88 55## Alias baseMean log2FoldChange lfcSE stat pvalue
## 80 PA14_00960 22.93951 -2.051543 0.5889029 -3.483669 4.945904e-04
## 90 PA14_01080 73.39200 -2.130515 0.4126581 -5.162907 2.431440e-07
## 232 PA14_03020 19.45471 2.077373 0.5814232 3.572910 3.530367e-04
## 334 PA14_04300 19736.03694 2.485906 0.2137388 11.630577 2.881430e-31
## 361 PA14_04650 13864.43974 2.045712 0.2117608 9.660485 4.437570e-22
## 499 PA14_06420 151.13934 -2.133427 0.2899808 -7.357133 1.879016e-13
## padj gene_id seqnames start end width strand source
## 80 2.461552e-03 gene1650995 chromosome 95402 95866 465 - PseudoCAP
## 90 3.527159e-06 gene1651015 chromosome 104835 105881 1047 + PseudoCAP
## 232 1.862733e-03 gene1651313 chromosome 266282 267136 855 - PseudoCAP
## 334 3.235270e-28 gene1651523 chromosome 384875 385207 333 - PseudoCAP
## 361 1.465442e-19 gene1651577 chromosome 413653 414192 540 - PseudoCAP
## 499 1.146608e-11 gene1651855 chromosome 564699 565457 759 + PseudoCAP
## type score phase Name Dbxref name Parent locus SK064_rpkm
## 80 gene NA 0 GeneID:4380425 PA14_00960 <NA> 4.215454
## 90 gene NA 0 GeneID:4383507 PA14_01080 <NA> 5.018643
## 232 gene NA 0 GeneID:4383655 PA14_03020 <NA> 1.925201
## 334 gene NA 0 GeneID:4383734 PA14_04300 <NA> 11.972453
## 361 gene NA 0 GeneID:4384279 pfpI <NA> 11.190724
## 499 gene NA 0 GeneID:4383861 PA14_06420 <NA> 6.436684
## SK065_rpkm SK066_rpkm SK070_rpkm SK071_rpkm SK072_rpkm SK064_cpm SK065_cpm
## 80 4.413025 4.304990 0.9728793 2.253514 2.951479 3.197449 3.384212
## 90 5.068786 4.300292 1.5480067 3.227785 2.485971 5.082906 5.133116
## 232 1.788204 1.166510 2.9155281 2.742662 3.105682 1.762222 1.631355
## 334 12.044165 12.233940 14.0352883 14.262723 14.093989 10.386766 10.458443
## 361 11.031055 11.484658 13.0182620 12.551957 12.957961 10.302281 10.142674
## 499 6.367097 6.140465 3.9768894 3.460016 4.049065 6.044135 5.974808
## SK066_cpm SK070_cpm SK071_cpm SK072_cpm SK064_rawcounts SK065_rawcounts
## 80 3.281914 0.5337433 1.460648 2.046829 41 54
## 90 4.363263 1.5919487 3.287117 2.540626 165 195
## 232 1.045584 2.7215930 2.552759 2.907825 12 12
## 334 10.648134 12.4490545 12.676464 12.507748 6711 8042
## 361 10.596118 12.1294414 11.663193 12.069147 6329 6460
## 499 5.749116 3.6078646 3.103226 3.678648 326 354
## SK066_rawcounts SK070_rawcounts SK071_rawcounts SK072_rawcounts locusId
## 80 41 2 9 13 2194652
## 90 92 9 45 20 2194662
## 232 5 25 25 27 2194811
## 334 7535 24976 33617 24169 2194915
## 361 7268 20012 16652 17832 2194942
## 499 248 50 39 49 2195081
## accession GI scaffoldId stop
## 80 YP_788236.1 116053799 4582 95402
## 90 YP_788246.1 116053809 4582 105881
## 232 YP_788395.1 116053957 4582 266282
## 334 YP_788499.1 116054056 4582 384875
## 361 YP_788526.1 116054083 4582 413653
## 499 YP_788665.1 116054221 4582 565457
## desc COG COGFun
## 80 putative lipoprotein (NCBI) COG3521 S
## 90 hypothetical protein (NCBI) COG3520 S
## 232 putative shikimate 5-dehydrogenase (NCBI) COG169 E
## 334 hypothetical protein (NCBI) COG3422 S
## 361 protease PfpI (NCBI) COG693 R
## 499 putative lactam utilization protein (NCBI) COG1540 R
## COGDesc
## 80 Uncharacterized protein conserved in bacteria
## 90 Uncharacterized protein conserved in bacteria
## 232 Shikimate 5-dehydrogenase
## 334 Uncharacterized conserved protein
## 361 Putative intracellular protease/amidase
## 499 Uncharacterized proteins, homologs of lactam utilization protein B
## TIGRFam
## 80 TIGR03352 type VI secretion lipoprotein, VC_A0113 family
## 90 TIGR03347 type VI secretion protein, VC_A0111 family
## 232
## 334
## 361 TIGR01382 intracellular protease, PfpI family
## 499
## TIGRRoles
## 80 Cellular processes:Pathogenesis
## 90 Cellular processes:Pathogenesis
## 232
## 334
## 361 Protein fate:Degradation of proteins, peptides, and glycopeptides
## 499
## GO EC ECDesc direction
## 80 down
## 90 down
## 232 GO:0005737,GO:0004764 1.1.1.25 Shikimate dehydrogenase. up
## 334 up
## 361 GO:0016798 3.2.-.- up
## 499 GO:0005975,GO:0003824 down10 PCA plot
## Warning in DESeqDataSet(se, design = design, ignoreRank): some variables in
## design formula are characters, converting to factors# Variance stabilizing transformations
vsd <- vst(vst, blind = TRUE)
# PCA plot
pcaData <- plotPCA(vsd, intgroup = c("exp_4", "batch", "condition"), returnData = TRUE)## using ntop=500 top features by variancepercentVar <- round(100 * attr(pcaData, "percentVar"))
# Calculate the maximum absolute value among PC1 and PC2
max_abs_val <- max(c(abs(min(pcaData$PC1)), abs(min(pcaData$PC2)), abs(max(pcaData$PC1)), abs(max(pcaData$PC2))))
axis_size <- 10
theme_set(theme_minimal(base_family = "Arial"))
pca <- ggplot(pcaData, aes(PC1, PC2, color = condition, shape = condition)) +
geom_point(size = 3) +
xlab(paste0("PC1: ", percentVar[1], "% variance")) +
ylab(paste0("PC2: ", percentVar[2], "% variance")) +
scale_colour_brewer(palette = "Dark2") +
scale_shape_manual(values = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)) +
coord_fixed(ratio = 1, xlim = c(-max_abs_val, max_abs_val), ylim = c(-max_abs_val, max_abs_val)) +
labs(color = "condition", shape = "condition") +
guides(shape = guide_legend(title = "Condition"), # Adjust size of shape legend
color = guide_legend(title = "Condition")) +
theme(
panel.grid = element_blank(),
panel.background = element_blank(),
axis.text = element_text(size = axis_size),
axis.ticks = element_line(linewidth = 0.2),
axis.line = element_line(linewidth = 0, color = "black"), # Set the axis line linewidth
panel.border = element_rect(fill = NA, linewidth = 0.5), # Set the panel border linewidth
legend.key = element_rect(fill = "white", color = NA) # Set legend key background to white and remove border
)
print(pca)
cairo_pdf(paste0(Sys.Date(),"_pca_",exp, "_", treatment,".vs.", control,".pdf", sep = ""))
pca
dev.off()## quartz_off_screen
## 211 Volcano plot
sequence <- c("treatment")
for (i in sequence) {
set.seed(0)
res <- results(dds_final, contrast = c("exp_4", i, "control"))
gp2 <- wes_palettes$Zissou1
key <- read.csv("../../raw-data/pa14_gene_key.csv", header = TRUE)
rownames(key) <- key$gene_id
res <- merge(key, as.data.frame(res), by=0)
res <- as.data.frame(res)
res$Alias <- rownames(res)
# Define gene list
alg_genes <- c(
"algA", "algF", "algJ", "algI", "algL", "algX", "algG", "algE",
"algK", "alg44", "alg8", "algD", "mucC", "mucB", "algU", "algW",
"algP", "algQ", "algR", "algZ", "algC", "algB"
)
logfc_cutoff
keyvals <- ifelse(
res$log2FoldChange <= -logfc_cutoff & res$padj <= pval_cutoff, gp2[1],
ifelse(res$log2FoldChange >= logfc_cutoff & res$padj <= pval_cutoff, gp2[5],
'lightgrey'))
keyvals[is.na(keyvals)] <- 'lightgrey'
names(keyvals)[keyvals == gp2[5]] <- 'Up-regulated'
names(keyvals)[keyvals == 'lightgrey'] <- 'NS'
names(keyvals)[keyvals == gp2[1]] <- 'Down-regulated'
volcano <- EnhancedVolcano(res,
lab = res[["name"]],
selectLab = alg_genes,
x = 'log2FoldChange',
y = 'padj',
xlab = bquote(~Log[2]~ 'fold change'),
pCutoff = pval_cutoff,
FCcutoff = logfc_cutoff,
cutoffLineType = "dashed",
pointSize = 1,
labSize = 3.0,
axisLabSize = 14,
labCol = 'black',
labFace = 'bold',
#col = c('grey', gp2[2], gp2[3], gp2[1]),
colCustom = keyvals,
colAlpha = 5/5,
legendPosition = 'bottom',
legendLabSize = 12,
legendIconSize = 4.0,
drawConnectors = TRUE,
arrowheads = FALSE,
widthConnectors = 0.2,
gridlines.major = FALSE,
borderWidth = 0.3,
gridlines.minor = FALSE,
title = paste0(treatment," vs ", control, sep = ""),
subtitle = "",
#caption = paste0("total = ", nrow(toptable), " variables"),
caption = paste0("Volcano plot with log2 FC of ", logfc_cutoff," and P value of ",pval_cutoff, "."),
max.overlaps = 30,
#ylim = c(0,10),
colConnectors = 'black') +
theme(axis.ticks=element_line(linewidth =0.2)) # +
# scale_x_continuous(breaks = c(-5, -2.5, 0, 2.5, 5), limits = c(-6,6))
print(volcano)
cairo_pdf(paste0(Sys.Date(),"_volcano_", exp, "_", treatment,".vs.", control,".pdf", sep = ""))
print(volcano)
dev.off()
}
12 MA plot
sequence <- c("treatment")
for (i in sequence) {
set.seed(0)
res <- results(dds_final, contrast = c("exp_4", i, "control"))
gp2 <- wes_palettes$Zissou1
res$gene_id <- rownames(res)
res <- as.data.frame(res)
head(res)
res <- full_join(key, as.data.frame(res), by="gene_id")
rownames(res) <- res$Alias
any(is.na(res)) # Check if there are any NAs
res <- res %>% filter(!is.na(baseMean))
ma <- ggmaplot(res,
fdr = pval_cutoff,
fc = 2^(logfc_cutoff),
size = 1,
palette = c((gp2[5]), (gp2[1])),
select.top.method = "fc",
# ylim = c(-15,15),
legend = "bottom",
title = paste0(treatment," vs ", control, sep = ""),
subtitle = paste0("Volcano plot with log2 FC of ", logfc_cutoff," and P value of ",pval_cutoff, "."),
font.label = c( 12),
label.rectangle = FALSE,
genenames = res$name,
font.legend = c(12),
font.main = c("bold", 14),
alpha = 1.5,
ggtheme = ggplot2::theme(panel.grid =element_blank(),
axis.text = element_text(size = 12),
axis.title = element_text(size = 14),
panel.background = element_blank(),
panel.border = element_rect(fill = NA)))
print(ma)
cairo_pdf(paste0(Sys.Date(),"_ma_", exp, "_", treatment,".vs.", control,".pdf"))
print(ma)
dev.off()
}
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