Last updated: 2018-04-10
Code version: 77d10d4
library(Biobase)We collected two types of data for each single cell sample: single-cell RNA-seq using C1 plates and FUCCI image intensity data.
Raw RNA-seq data: data/eset-raw.rds
FUCCI intensity data: data/intensity.rds
Filtered RNA-seq data: output/gene-filtering.Rmd/eset-gene-filtering.rds
Final data combining filtered intensity and RNA-seq, including 12792 genes and 990 samples: data/eset-final.rds
Combined intensity data are stored in data/intensity.rds. These include samples that were identified to have a single nuclei .
Data generated by [combine-intensity-data.R][data/combine-intensity-data.R]. Combining image analysis output stored in /project2/gilad/fucci-seq/intensities_stats/ into one data.frame and computes summary statistics, including background-corrected RFP and GFP intensity measures.
ints <- readRDS(file="../data/intensity.rds")
colnames(ints) [1] "plate" "image"
[3] "size" "perimeter"
[5] "eccentricity" "rfp.fore.zoom.mean"
[7] "rfp.fore.zoom.median" "gfp.fore.zoom.mean"
[9] "gfp.fore.zoom.median" "dapi.fore.zoom.mean"
[11] "dapi.fore.zoom.median" "rfp.back.zoom.mean"
[13] "rfp.back.zoom.median" "gfp.back.zoom.mean"
[15] "gfp.back.zoom.median" "dapi.back.zoom.mean"
[17] "dapi.back.zoom.median" "rfp.mean.log10sum"
[19] "gfp.mean.log10sum" "dapi.mean.log10sum"
[21] "rfp.median.log10sum" "gfp.median.log10sum"
[23] "dapi.median.log10sum" "unique"
[25] "chip_id" Raw data from each C1 plate are stored separatley in data/eset/ by experiment (batch) ID.
Raw data combining C1 plate are stored in data/eset-raw.rds.
Filtered raw data excluding low-quality sequencing samples and genes that are lowly expressed or overly expressed are stored in output/gene-filtering.Rmd/eset-gene-filtering.rds (both eSet and CPM data.table).
eset_filtered <- readRDS("../output/gene-filtering.Rmd/eset-gene-filtering.rds")We match labels in the intensity data and in the sequencing data. 990 samples are quantified in both datasets.
pdata <- pData(eset_filtered)
pdata$unique <- paste(pdata$image_individual, sprintf("%05d", pdata$image_label), sep="_")
sample_include_bothdata <- intersect(ints$unique, pdata$unique)
length(sample_include_bothdata)[1] 990Make a combined eSet object include both FUCCI intensity data and RNA-seq data.
ints_combo <- ints[which(ints$unique %in% sample_include_bothdata), ]
eset_combo <- new("ExpressionSet",
exprs = exprs(eset_filtered)[,which(pdata$unique %in% sample_include_bothdata)],
phenoData = phenoData(eset_filtered)[which(pdata$unique %in% sample_include_bothdata), ],
featureData = featureData(eset_filtered))
pdata_combo <- pData(eset_combo)
pdata_combo$unique <- paste(pdata_combo$image_individual,
sprintf("%05d", pdata_combo$image_label), sep="_")
all.equal(ints_combo$unique, pdata_combo$unique)Import estimated cell time. Computed using cell times adjusted for plate effect ../output/images-normalize-anova.Rmd/pdata.adj.rds. See ../docs/npreg.html for methods.
theta <- readRDS(file = "../output/npreg.Rmd/theta.rds")Make variable labels.
pdata_table <- rbind(varMetadata(phenoData(eset_combo)),
"mCherry background-corrected intensity (log10sum)",
"EGFP background-corrected intensity (log10sum)",
"DAPI background-corrected intensity (log10sum)",
"nucleus size",
"nucleus perimeter",
"nucleus eccentricity",
"cell time")
rownames(pdata_table) <- c(rownames(varMetadata(phenoData(eset_combo))),
"rfp.median.log10sum",
"gfp.median.log10sum",
"dapi.median.log10sum",
"size", "perimeter", "eccentricity",
"theta")
phenoData(eset_combo) <- new("AnnotatedDataFrame",
data = data.frame(pData(eset_combo),
rfp.median.log10sum=ints_combo$rfp.median.log10sum,
gfp.median.log10sum=ints_combo$gfp.median.log10sum, dapi.median.log10sum=ints_combo$dapi.median.log10sum,
size=ints_combo$size,
perimeter=ints_combo$perimeter,
eccentricity=ints_combo$eccentricity,
theta=theta),
varMetadata = pdata_table)
saveRDS(eset_combo, file = "../data/eset-final.rds")We store feature-level (gene) read count and molecule count in expressionSet (data/eset) objects, which also contain sample metadata (e.g., assigned indivdual ID, cDNA concentraion) and quality filtering criteria (e.g., number of reads mapped to FUCCI transgenes, ERCC conversion rate). Data from different C1 plates are stored in separate eset objects:
To combine eset objects from the different C1 plates:
eset <- Reduce(combine, Map(readRDS, Sys.glob("data/eset/*.rds")))
To access data stored in expressionSet:
exprs(eset): access count data, 20,421 features by 1,536 single cell samples.
pData(eset): access sample metadata. Returns data.frame of 1,536 samples by 43 labels. Use varMetadata(phenoData(eset)) to view label descriptions.
fData(eset): access feature metadata. Returns data.frame of 20,421 features by 6 labels. Use varMetadata(featureData(eset)) to view label descriptions.
varMetadata(phenoData(eset)): view the sample metadata labels.
varMetadata(featureData(eset)): view the feature (gene) metadata labels.
R version 3.4.1 (2017-06-30)
Platform: x86_64-redhat-linux-gnu (64-bit)
Running under: Scientific Linux 7.2 (Nitrogen)
Matrix products: default
BLAS/LAPACK: /usr/lib64/R/lib/libRblas.so
locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] parallel stats graphics grDevices utils datasets methods
[8] base
other attached packages:
[1] Biobase_2.38.0 BiocGenerics_0.24.0
loaded via a namespace (and not attached):
[1] Rcpp_0.12.16 digest_0.6.15 rprojroot_1.3-2 backports_1.1.2
[5] git2r_0.21.0 magrittr_1.5 evaluate_0.10.1 stringi_1.1.7
[9] rmarkdown_1.9 tools_3.4.1 stringr_1.3.0 yaml_2.1.18
[13] compiler_3.4.1 htmltools_0.3.6 knitr_1.20 This R Markdown site was created with workflowr