1 Introduction

ASOC generated this report includes stats and diagnostic plots of the Xenium In Situ data.

2 Prerequisite

2.1 Environment

OS: macOS Sequoia 15.0
Platform: aarch64-apple-darwin20 (64-bit)
Software: R (v4.4.1)
Pakcages: arrow (v17.0.0.1), dplyr (v1.1.4), ggplot2 (v3.5.1), ggrepel (v0.9.5), ggridges (v0.5.6), HDF5Array (v1.32.1), outliers (v0.15), presto (v1.0.0), reshape2 (v1.4.4), scMerge (v1.20.0), SingleCellExperiment (v1.26.0), stringr (v1.5.1), yaml (v2.3.10)

2.2 Xenium In Situ

Slide ID: 0010348
Preparation method: FFPE
Sample description: Six female mouse brains
- 5xFAD transgenic (TG) mice (greens, even numbers), N=3:
  - TG2F (Region_2) ◼
  - TG3F (Region_4) ◼
  - TG4F (Region_6) ◼
- Wild-type (WT) mice (purples, odd numbers), N=3:
  - WT1F (Region_1) ◼
  - WT2F (Region_3) ◼
  - WT3F (Region_5) ◼
Number of panel genes: 247
Xenium Ranger analysis summary
Files: feature.tsv.gz, cell_feature_matrix.h5, cells.parquet, metrics_summary.csv, and transcripts.parquet (see below)

WORKING_DIRECTORY
└── output-INSTRUMENT__SLIDE__REGION__DATE__TIME
    ├── cell_feature_matrix
    │   └── features.tsv.gz
    ├── cell_feature_matrix.h5
    ├── cells.parquet
    ├── metrics_summary.csv
    └── transcripts.parquet

3 Preprocessing/QC

3.1 Preparation

Set working directory to load files

# Do not run
baseDir <- "WORKING_DIRECTORY" # where the output-* folders are located
setwd(baseDir)
outDir <- file.path(baseDir, "Results")

library(arrow)
library(dplyr)
library(ggplot2)
library(ggrepel)
library(ggridges)
library(HDF5Array)
library(outliers)
library(presto)
library(reshape2)
library(scMerge)
library(SingleCellExperiment)
library(stringr)

3.2 Load Xenium data (Step 1)

Import H5 and Parquet files
Source: https://www.10xgenomics.com/support/software/cell-ranger/latest/analysis/outputs/cr-outputs-h5-matrices
Source: https://www.10xgenomics.com/support/software/xenium-onboard-analysis/latest/advanced/example-code

outputFolders <- list.files(path = baseDir, pattern = "output")
outputFolders

## [1] "output-XETG00116__0010348__TG2F__20240427__002555"
## [2] "output-XETG00116__0010348__TG3F__20240427__002555"
## [3] "output-XETG00116__0010348__TG4F__20240427__002555"
## [4] "output-XETG00116__0010348__WT1F__20240427__002555"
## [5] "output-XETG00116__0010348__WT2F__20240427__002555"
## [6] "output-XETG00116__0010348__WT3F__20240427__002555"

# Summary
summaryL <- lapply(seq_along(outputFolders), function(idx) {
        outputFolder <- outputFolders[idx]
        summary <- read.csv(file.path(baseDir, outputFolder, "metrics_summary.csv"))
        return(summary)
})
summary <- Reduce(rbind, summaryL)

# Cells
cellsL <- lapply(seq_along(outputFolders), function(idx) {
        outputFolder <- outputFolders[idx]
        cells <- arrow::read_parquet(file.path(baseDir, outputFolder, "cells.parquet"), as_data_frame = TRUE)
        cells$region <- sapply(str_split(outputFolder, "__"), "[[", 3)
        return(cells)
})
cells <- Reduce(rbind, cellsL)
cells$region <- factor(cells$region, levels = regionNames)

# Transcripts
txL <- lapply(seq_along(outputFolders), function(idx) {
        outputFolder <- outputFolders[idx]
        tx <- arrow::read_parquet(file.path(baseDir, outputFolder, "transcripts.parquet"), as_data_frame = TRUE)
        return(tx)
})
names(txL) <- sapply(str_split(outputFolders, "__"), "[[", 3)
txL <- txL[regionNames]

# Expression profile
exprL <- lapply(seq_along(outputFolders), function(idx) {
        outputFolder <- outputFolders[idx]
        mat <- HDF5Array::TENxMatrix(file.path(baseDir, outputFolder, "cell_feature_matrix.h5"), "matrix")
        return(mat)
})
names(exprL) <- sapply(str_split(outputFolders, "__"), "[[", 3)
exprL <- exprL[regionNames]

# Feature annotation
featuresAnnot <- read.delim(file.path(baseDir, outputFolders[1], "cell_feature_matrix", "features.tsv.gz"), header = FALSE, stringsAsFactors = FALSE)

3.3 Stats at a glance

Region	Region_area	Cell_area	Cells_detected	High_qual_Tx	Tx_per_100um2	FOVs
TG2F	28462857	8833587	53020	11305322	127.9811	65
TG3F	23250723	8057483	51515	9050443	112.3235	53
TG4F	21498027	7171907	45791	7632048	106.4159	49
WT1F	28483607	9199148	54878	12407959	134.8816	65
WT2F	28034067	7882022	47444	10679458	135.4914	64
WT3F	23244105	7663646	50510	8058815	105.1564	53

3.4 Exclude unassigned Tx (Step 2)

The following capture is an example case to show Tx that are unassigned

txAssignedL <- lapply(seq_along(txL), function(idx) {
        tx <- txL[[idx]]
        return(tx[which(tx$cell_id != "UNASSIGNED"), ])
})
names(txAssignedL) <- regionNames

Model	Total	AssignedTx	UnAssignedTx	Proportion
TG2F	13870843	9774591	4096252	0.295
TG3F	11414778	8855622	2559156	0.224
TG4F	9643010	7360260	2282750	0.237
WT1F	14980357	10621294	4359063	0.291
WT2F	13171948	9404299	3767649	0.286
WT3F	10177480	7965262	2212218	0.217
Total	73258416	53981328	19277088	0.263

3.5 QC (Step 3)

3.5.1 Cell area and detected Tx

The following violin plots show the distribution of the cell area or detected Tx across regions
- X-axis: model
- Y-axis: area in micrometer square or total detected Tx
- Dot: cell

ggplot(data = cells, aes(x = region, y = log10(cell_area + 1), fill = region)) +
        geom_violin(position = dodge, size = 0) +
        geom_boxplot(width = 0.1, position = dodge, fill = "white") +
        scale_fill_manual(values = regionCols) +
        labs(
                x = "",
                y = "Cell area, log10"
        ) +
        theme_bw() +
        theme(
                axis.line = element_line(colour = "black"),
                panel.grid.major = element_blank(),
                panel.grid.minor = element_blank(),
                panel.border = element_blank(),
                panel.background = element_blank(),
                axis.text.x = element_text(angle = 0, vjust = 0, hjust = 0.5),
                legend.position = "none",
                text = element_text(size = 12)
        )

ggplot(data = cells, aes(x = region, y = log10(transcript_counts + 1), fill = region)) +
        geom_violin(position = dodge, size = 0) +
        geom_boxplot(width = 0.1, position = dodge, fill = "white") +
        scale_fill_manual(values = regionCols) +
        labs(
                x = "",
                y = "Tx counts, log10"
        ) +
        theme_bw() +
        theme(
                axis.line = element_line(colour = "black"),
                panel.grid.major = element_blank(),
                panel.grid.minor = element_blank(),
                panel.border = element_blank(),
                panel.background = element_blank(),
                axis.text.x = element_text(angle = 0, vjust = 0, hjust = 0.5),
                legend.position = "none",
                text = element_text(size = 12)
        )

3.5.2 Detect ouliters in terms of area

Detect under segmentated (when more than one cells segmented as one) cells using Grubb’s test
A given area of the cell is an outlier when Grubb’s test p-value is less than 0.05 in a region
Note that a centroid diameter, or cell diameter, is preset in Xenium per tissue type

Model	Total_cells	Excluded	Remainders
TG2F	53020	11	53009
TG3F	51515	5	51510
TG4F	45791	2	45789
WT1F	54878	5	54873
WT2F	47444	3	47441
WT3F	50510	2	50508
Total	303158	28	303130

for (region in regionNames) {
        subCells <- cells[which(cells$region == region), ]
        subCells$cell_area <- round(subCells$cell_area, 3)
        testRes <- grubbsTestRec(subCells$cell_area)
        areaThreshold <- max(testRes$Area[!testRes$Outlier])
        
        cat("#### ", region, "\n")
        cat("\n")
        plot(density(testRes$Area), sub="Cell area, µm2", main="")
        abline(v = areaThreshold, col="red", lty=2)
        cat("\n\n")
}

3.5.2.1 TG2F

3.5.2.2 TG3F

3.5.2.3 TG4F

3.5.2.4 WT1F

3.5.2.5 WT2F

3.5.2.6 WT3F

3.5.3 Detected Tx across cells

Number of detected Tx across cells at the FOV level
In other words, some Tx detected only in one cell, while others detected in many cells
The following ridge plot shows the distribution of the detected Tx in each FOV
- X-axis: number of cells that captures a given Tx
- Y-axis: density

txDfL <- lapply(seq_along(txAssignedL), function(idx) {
        txName <- names(txAssignedL)[idx]
        txAssigned <- txAssignedL[[idx]]
        txDf <- as.data.frame(txAssigned %>% group_by(feature_name, fov_name) %>% dplyr::count() %>% dplyr::rename(cells = n))
        txDf$model <- txName
        return(txDf)
})
txDf <- Reduce(rbind, txDfL)
head(txDf)

##    feature_name fov_name cells model
## 1 2010300C02Rik      E12  3789  TG2F
## 2 2010300C02Rik      E13  5266  TG2F
## 3 2010300C02Rik      E14  6676  TG2F
## 4 2010300C02Rik      E15  4288  TG2F
## 5 2010300C02Rik      E16  2035  TG2F
## 6 2010300C02Rik      E17   339  TG2F

for (region in regionNames) {
        cat("#### ", region, "\n")
        cat("\n")
        regDf <- txDf[which(txDf$model == region), ]
        rp <- ridgePlot(df = regDf, x = "cells", y = "fov_name", title = region, xLbl = "nCells", yLbl = "FOV", scaleTrans = "log10")
        print(rp)
        cat("\n\n")
}

3.5.3.1 TG2F

3.5.3.2 TG3F

3.5.3.3 TG4F

3.5.3.4 WT1F

3.5.3.5 WT2F

3.5.3.6 WT3F

3.5.4 Count and Features per cell

Exclude cells with few (<10) Tx detected
The following scatter plots show the relationship between nCount and nFeature in linear (L) or log-scale (R)
- X-axis: nCount (total number of detected Tx within a cell, depth)
- Y-axis: nFeature (number of detected Tx in each cell, coverage)
- Dot: cell

Model	Total_cells	Large_cell_area	Low_nCount	Remainders
TG2F	53020	11	99	52910
TG3F	51515	5	207	51303
TG4F	45791	2	414	45375
WT1F	54878	5	93	54780
WT2F	47444	3	113	47328
WT3F	50510	2	386	50122
Total	303158	28	1312	301818

for (idx in seq_along(regionNames)) {
        exprName <- names(exprL)[idx]
        expr <- exprL[[idx]]
        panelGenesIdx <- which(str_detect(rownames(expr), "ENS"))
        expr <- expr[panelGenesIdx, ]

        cellDf <- data.frame(
                nCount = apply(as.matrix(expr), 2, sum),
                nFeature = apply(as.matrix(expr), 2, function(x) length(which(x > 0)))
        )
        cellDf <- densityColors(df = cellDf, cols = heatCols)

        cellDf$criteria <- "Include"
        cellDf$criteria[which(cellDf$nCount < nCountThre)] <- "Exclude"

        cellDfInc <- cellDf[which(cellDf$criteria == "Include"), ]
        cellDfExc <- cellDf[which(cellDf$criteria == "Exclude"), ]

        cat("#### ", exprName, "\n")
        cat("\n")
        par(mfrow = c(2, 2))
        plot(cellDf$nCount, cellDf$nFeature, col = cellDf$Col, xlab = "nCount", ylab = "nFeature", pch = 20, xlim = c(0, nCountMax), ylim = c(0, nFeatureMax))
        plot(log10(cellDf$nCount + 1), log10(cellDf$nFeature + 1), col = cellDf$Col, xlab = "nCount, log10", ylab = "nFeature, log10", pch = 20, xlim = c(0, log10(nCountMax)), ylim = c(0, log10(nFeatureMax)))
        plot(cellDfInc$nCount, cellDfInc$nFeature, col = "grey60", xlab = "nCount", ylab = "nFeature", pch = 20, xlim = c(0, nCountMax), ylim = c(0, max(cellDfInc$nFeature)))
        points(cellDfExc$nCount, cellDfExc$nFeature, col = "red", pch = 20)
        legend("topleft", legend = c("Include", "Exclude"), fill = c("grey60", "red"), bty = "n")
        plot(log10(cellDfInc$nCount + 1), log10(cellDfInc$nFeature + 1), col = "grey60", xlab = "nCount, log10", ylab = "nFeature, log10", pch = 20, xlim = c(0, log10(nCountMax + 1)), ylim = c(0, log10(nFeatureMax)))
        points(log10(cellDfExc$nCount + 1), log10(cellDfExc$nFeature + 1), col = "red", pch = 20)
        legend("topleft", legend = c("Include", "Exclude"), fill = c("grey60", "red"), bty = "n")
        cat("\n\n")
}

3.5.4.1 TG2F

3.5.4.2 TG3F

3.5.4.3 TG4F

3.5.4.4 WT1F

3.5.4.5 WT2F

3.5.4.6 WT3F

3.5.5 Background signals

Exclude a cell when a background signal (proportion) is greater than 0.05
- Proportion = Number of negative probe detected / Number of (Tx + Neg probe) detected
Negative control probes are probes that exist in the panels but target non-biological sequences.
Negative control codewords are codewords in the codebook that do not have any probes matching that code.
Source: https://www.10xgenomics.com/support/software/xenium-onboard-analysis/latest/algorithms-overview/xoa-algorithms

Model	Total_cells	Large_cell_area	Low_nCount	High_background	Remainders
TG2F	53020	11	99	3	52907
TG3F	51515	5	207	32	51271
TG4F	45791	2	414	88	45287
WT1F	54878	5	93	2	54778
WT2F	47444	3	113	2	47326
WT3F	50510	2	386	94	50028
Total	303158	28	1312	221	301597

for (idx in seq_along(regionNames)) {
        exprName <- names(exprL)[idx]
        expr <- exprL[[idx]]

        panelGenesIdx <- which(str_detect(rownames(expr), "^ENS"))
        negProbesIdx <- which(str_detect(rownames(expr), "^NegControlProbe"))
        # negCodewordsIdx <- which(str_detect(rownames(expr), "^NegControlCodeword"))

        signalDf <- data.frame(
                CellID = colnames(expr),
                Genes = apply(as.matrix(expr[panelGenesIdx, ]), 2, sum),
                negProbes = apply(as.matrix(expr[negProbesIdx, ]), 2, sum)
                # negCodewords = apply(expr[negCodewordsIdx,], 2, sum)
        )
        signalDf <- signalDf[which(signalDf$negProbes != 0), ]
        signalDf$Prop <- signalDf$negProbes / (signalDf$Genes + signalDf$negProbes)

        cat("#### ", exprName, "\n")
        cat("\n")
        hist(signalDf$Prop, breaks = seq(0, max(signalDf$Prop) + 0.01, 0.01), xlab = "# Neg probes / Total detected molecules", sub = paste0(nrow(signalDf), " cells that capture at least one Neg Probe"), main = exprName)
        abline(v = negProbePropThre, col = "red", lty = 2)
        cat("\n\n")
}

3.5.5.1 TG2F

3.5.5.2 TG3F

3.5.5.3 TG4F

3.5.5.4 WT1F

3.5.5.5 WT2F

3.5.5.6 WT3F

3.5.6 Tx Quality and Number of Tx

10X Genomics recommend to include Tx quality that is greater than 20 and the threshold in this study is 20
- 20 (99%, an error rate of 1 in 100)
- 30 (99.9%, an error rate of 1 in 1000)
- 40 (99.99%, an error rate of 1 in 10000)
Low qual Tx are likely the negative controls
QC purposes across models/runs/cohorts

qvDf <- c()
for (idx in seq_along(regionNames)) {
        txName <- names(txAssignedL)[idx]
        txAssigned <- txAssignedL[[idx]]

        buff <- data.frame(cut(txAssigned$qv, breaks = c(0, 20, 30, 40)) %>% table)
        buff$Model <- txName

        qvDf <- rbind(qvDf, buff)
}
colnames(qvDf) <- c("Quality_score", "Freq", "Model")
qvDf$Model <- factor(qvDf$Model, levels=regionNames)
knitr::kable(reshape2::dcast(qvDf, Model ~ Quality_score, value.var = c("Freq")))

Model	(0,20]	(20,30]	(30,40]
TG2F	1786996	639617	7347978
TG3F	1779629	677997	6397996
TG4F	1475614	1038204	4846442
WT1F	1889126	719979	8012189
WT2F	1811919	668407	6923973
WT3F	1609489	696525	5659248

3.5.7 QC Summary

Number of cells after QC for the downstream analysis
Analysis ready expression profiles (before and after QC)

exprL <- lapply(seq_along(exprL), function(idx) {
        expr <- exprL[[idx]]
        panelGenesIdx <- which(str_detect(rownames(expr), "ENS"))
        expr <- expr[panelGenesIdx, ]
        return(expr)
})
names(exprL) <- regionNames

arExprL <- lapply(seq_along(exprL), function(idx) { # analysis-ready expression profiles
        exprName <- names(exprL)[idx]
        expr <- exprL[[idx]]
        exclude <- excludeCells[which(excludeCells$Model == exprName),]

        if (any(colnames(expr) %in% exclude$Cells)) {
                expr <- expr[, -which(colnames(expr) %in% exclude$Cells)]
        }
        return(expr)
})
names(arExprL) <- regionNames
saveRDS(arExprL, file.path(outDir, "01_analysisReady_expression.RDS"))

Model	Total_cells	Excluded_cells	Proportion_exc	Remainders
TG2F	53020	113	0.002	52907
TG3F	51515	241	0.005	51274
TG4F	45791	494	0.011	45297
WT1F	54878	100	0.002	54778
WT2F	47444	118	0.002	47326
WT3F	50510	471	0.009	50039
Total	303158	1537	0.005	301621

4 Research reproducibility

Parameters used in this report

Step	Parameter	Threshold	Description
QC	nCount	10	Minimum number of detected Tx within a cell
	negProbe proportion	0.05	Maximum proportion of the background signal
	quality score	20	Minimum Tx Phred-like quality score

sessionInfo()

## R version 4.4.1 (2024-06-14)
## Platform: aarch64-apple-darwin20
## Running under: macOS 15.0
## 
## Matrix products: default
## BLAS:   /Library/Frameworks/R.framework/Versions/4.4-arm64/Resources/lib/libRblas.0.dylib 
## LAPACK: /Library/Frameworks/R.framework/Versions/4.4-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.12.0
## 
## locale:
## [1] en_CA.UTF-8/en_CA.UTF-8/en_CA.UTF-8/C/en_CA.UTF-8/en_CA.UTF-8
## 
## time zone: America/Edmonton
## tzcode source: internal
## 
## attached base packages:
## [1] stats4    stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
##  [1] stringr_1.5.1               SingleCellExperiment_1.26.0
##  [3] SummarizedExperiment_1.34.0 Biobase_2.64.0             
##  [5] GenomicRanges_1.56.1        GenomeInfoDb_1.40.1        
##  [7] scMerge_1.20.0              reshape2_1.4.4             
##  [9] presto_1.0.0                data.table_1.16.0          
## [11] Rcpp_1.0.13                 outliers_0.15              
## [13] HDF5Array_1.32.1            rhdf5_2.48.0               
## [15] DelayedArray_0.30.1         SparseArray_1.4.8          
## [17] S4Arrays_1.4.1              abind_1.4-8                
## [19] IRanges_2.38.1              S4Vectors_0.42.1           
## [21] MatrixGenerics_1.16.0       matrixStats_1.4.1          
## [23] BiocGenerics_0.50.0         Matrix_1.7-0               
## [25] ggridges_0.5.6              ggrepel_0.9.6              
## [27] ggplot2_3.5.1               dplyr_1.1.4                
## [29] arrow_17.0.0.1             
## 
## loaded via a namespace (and not attached):
##   [1] RColorBrewer_1.1-3        rstudioapi_0.16.0        
##   [3] jsonlite_1.8.9            magrittr_2.0.3           
##   [5] ggbeeswarm_0.7.2          farver_2.1.2             
##   [7] rmarkdown_2.28            zlibbioc_1.50.0          
##   [9] vctrs_0.6.5               DelayedMatrixStats_1.26.0
##  [11] base64enc_0.1-3           htmltools_0.5.8.1        
##  [13] BiocNeighbors_1.22.0      Rhdf5lib_1.26.0          
##  [15] Formula_1.2-5             sass_0.4.9               
##  [17] StanHeaders_2.32.10       reldist_1.7-2            
##  [19] KernSmooth_2.23-24        bslib_0.8.0              
##  [21] htmlwidgets_1.6.4         plyr_1.8.9               
##  [23] cachem_1.1.0              ResidualMatrix_1.14.1    
##  [25] sfsmisc_1.1-19            igraph_2.0.3             
##  [27] startupmsg_0.9.7          lifecycle_1.0.4          
##  [29] pkgconfig_2.0.3           M3Drop_1.30.0            
##  [31] rsvd_1.0.5                R6_2.5.1                 
##  [33] fastmap_1.2.0             GenomeInfoDbData_1.2.12  
##  [35] digest_0.6.37             numDeriv_2016.8-1.1      
##  [37] colorspace_2.1-1          scater_1.32.1            
##  [39] dqrng_0.4.1               irlba_2.3.5.1            
##  [41] Hmisc_5.1-3               beachmat_2.20.0          
##  [43] labeling_0.4.3            fansi_1.0.6              
##  [45] httr_1.4.7                mgcv_1.9-1               
##  [47] compiler_4.4.1            bit64_4.0.5              
##  [49] withr_3.0.1               htmlTable_2.4.3          
##  [51] backports_1.5.0           inline_0.3.19            
##  [53] BiocParallel_1.38.0       viridis_0.6.5            
##  [55] highr_0.11                QuickJSR_1.3.1           
##  [57] pkgbuild_1.4.4            gplots_3.1.3.1           
##  [59] MASS_7.3-61               proxyC_0.4.1             
##  [61] bluster_1.14.0            gtools_3.9.5             
##  [63] caTools_1.18.3            loo_2.8.0                
##  [65] distr_2.9.5               tools_4.4.1              
##  [67] vipor_0.4.7               foreign_0.8-87           
##  [69] beeswarm_0.4.0            nnet_7.3-19              
##  [71] glue_1.7.0                batchelor_1.20.0         
##  [73] cvTools_0.3.3             nlme_3.1-166             
##  [75] rhdf5filters_1.16.0       grid_4.4.1               
##  [77] checkmate_2.3.2           cluster_2.1.6            
##  [79] generics_0.1.3            gtable_0.3.5             
##  [81] metapod_1.12.0            BiocSingular_1.20.0      
##  [83] ScaledMatrix_1.12.0       utf8_1.2.4               
##  [85] XVector_0.44.0            pillar_1.9.0             
##  [87] limma_3.60.4              robustbase_0.99-4        
##  [89] splines_4.4.1             lattice_0.22-6           
##  [91] densEstBayes_1.0-2.2      bit_4.0.5                
##  [93] ruv_0.9.7.1               tidyselect_1.2.1         
##  [95] locfit_1.5-9.10           scuttle_1.14.0           
##  [97] knitr_1.48                gridExtra_2.3            
##  [99] edgeR_4.2.1               xfun_0.47                
## [101] statmod_1.5.0             DEoptimR_1.1-3           
## [103] rstan_2.32.6              stringi_1.8.4            
## [105] UCSC.utils_1.0.0          yaml_2.3.10              
## [107] evaluate_1.0.0            codetools_0.2-20         
## [109] bbmle_1.0.25.1            tibble_3.2.1             
## [111] cli_3.6.3                 RcppParallel_5.1.9       
## [113] rpart_4.1.23              munsell_0.5.1            
## [115] jquerylib_0.1.4           bdsmatrix_1.3-7          
## [117] parallel_4.4.1            rstantools_2.4.0         
## [119] assertthat_0.2.1          scran_1.32.0             
## [121] sparseMatrixStats_1.16.0  bitops_1.0-8             
## [123] viridisLite_0.4.2         mvtnorm_1.3-1            
## [125] scales_1.3.0              purrr_1.0.2              
## [127] crayon_1.5.3              rlang_1.1.4

Spatial Transcriptomics at Single-cell Level

Heewon Seo

1 Introduction

2 Prerequisite

2.1 Environment

2.2 Xenium In Situ

3 Preprocessing/QC

3.1 Preparation

3.2 Load Xenium data (Step 1)

3.3 Stats at a glance

3.4 Exclude unassigned Tx (Step 2)

3.5 QC (Step 3)

3.5.1 Cell area and detected Tx

3.5.2 Detect ouliters in terms of area

3.5.2.1 TG2F

3.5.2.2 TG3F

3.5.2.3 TG4F

3.5.2.4 WT1F

3.5.2.5 WT2F

3.5.2.6 WT3F

3.5.3 Detected Tx across cells

3.5.3.1 TG2F

3.5.3.2 TG3F

3.5.3.3 TG4F

3.5.3.4 WT1F

3.5.3.5 WT2F

3.5.3.6 WT3F

3.5.4 Count and Features per cell

3.5.4.1 TG2F

3.5.4.2 TG3F

3.5.4.3 TG4F

3.5.4.4 WT1F

3.5.4.5 WT2F

3.5.4.6 WT3F

3.5.5 Background signals

3.5.5.1 TG2F

3.5.5.2 TG3F

3.5.5.3 TG4F

3.5.5.4 WT1F

3.5.5.5 WT2F

3.5.5.6 WT3F

3.5.6 Tx Quality and Number of Tx

3.5.7 QC Summary

4 Research reproducibility