PRECAST: Human Breast Cancer Data Analysis

This vignette introduces the PRECAST workflow for the analysis of integrating multiple spatial transcriptomics datasets. The workflow consists of three steps

  • Independent preprocessing and model setting
  • Probabilistic embedding, clustering and alignment using PRECAST model
  • Downstream analysis (i.e. visualization of clusters and embeddings, combined differential expression analysis)

We demonstrate the use of PRECAST to two sliced human breast cancer Visium data that are here, which can be downloaded to the current working path by the following command:

githubURL <- "https://github.com/feiyoung/PRECAST/blob/main/vignettes_data/bc2.rda?raw=true"
download.file(githubURL, "bc2.rda", mode = "wb")

Then load to R

load("bc2.rda")
Download data from 10X: another method to access data

This data is also available at 10X genomics data website:

Users require the two folders for each dataset: spatial and filtered_feature_bc_matrix. Then the data can be read by the following commond.

dir.file <- "Section"  ## the folders Section1 and Section2, and each includes two folders spatial and filtered_feature_bc_matrix
seuList <- list()
for (r in 1:2) {
    message("r = ", r)
    seuList[[r]] <- DR.SC::read10XVisium(paste0(dir.file, r))
}
bc2 <- seuList

The package can be loaded with the command:

library(PRECAST)
library(Seurat)

View human breast cancer Visium data from DataPRECAST

bc2  ## a list including two Seurat object

Check the content in bc2

head(bc2[[1]])

Create a PRECASTObject object

We show how to create a PRECASTObject object step by step. First, we create a Seurat list object using the count matrix and meta data of each data batch. Although bc2 is a prepared Seurat list object, we re-create it to show the details of the Seurat list object. At the same time, check the meta data that must include the spatial coordinates named “row” and “col”, respectively. If the names are not, they are required to rename them.

## Get the gene-by-spot read count matrices countList <- lapply(bc2, function(x)
## x[['RNA']]@counts)
countList <- lapply(bc2, function(x) {
    assay <- DefaultAssay(x)
    GetAssayData(x, assay = assay, slot = "counts")

})

M <- length(countList)
## Get the meta data of each spot for each data batch
metadataList <- lapply(bc2, function(x) x@meta.data)

for (r in 1:M) {
    meta_data <- metadataList[[r]]
    all(c("row", "col") %in% colnames(meta_data))  ## the names are correct!
    head(meta_data[, c("row", "col")])
}


## ensure the row.names of metadata in metaList are the same as that of colnames count matrix
## in countList

for (r in 1:M) {
    row.names(metadataList[[r]]) <- colnames(countList[[r]])
}


## Create the Seurat list object

seuList <- list()
for (r in 1:M) {
    seuList[[r]] <- CreateSeuratObject(counts = countList[[r]], meta.data = metadataList[[r]], project = "BreastCancerPRECAST")
}

bc2 <- seuList
rm(seuList)
head(meta_data[, c("row", "col")])

Prepare the PRECASTObject with preprocessing step.

Next, we use CreatePRECASTObject() to create a PRECASTObject based on the Seurat list object bc2. This function will do three things:

    1. Filter low-quality spots and genes, controlled by the arguments premin.features and premin.spots, respectively; the spots are retained in raw data (bc2) with at least premin.features number of nonzero-count features (genes), and the genes are retained in raw data (bc2) with at least premin.spots number of spots. To ease presentation, we denote the filtered Seurat list object as bc2_filter1.
    1. Select the top 2,000 variable genes (by setting gene.number=2000) for each data batch using FindSVGs() function in DR.SC package for spatially variable genes or FindVariableFeatures() function in Seurat package for highly variable genes. Next, we prioritized genes based on the number of times they were selected as variable genes in all samples and chose the top 2,000 genes. Then denote the Seurat list object as bc2_filter2, where only 2,000 genes are retained.
    1. Conduct strict quality control for bc2_filter2 by filtering spots and genes, controlled by the arguments postmin.features and postmin.spots, respectively; the spots are retained with at least post.features nonzero counts across genes; the features (genes) are retained with at least postmin.spots number of nonzero-count spots. Usually, no genes are filltered because these genes are variable genes.

If the argument customGenelist is not NULL, then this function only does (3) not (1) and (2). User can retain the raw seurat list object by setting rawData.preserve = TRUE.

## Create PRECASTObject.
set.seed(2022)
PRECASTObj <- CreatePRECASTObject(bc2, project = "BC2", gene.number = 2000, selectGenesMethod = "SPARK-X",
    premin.spots = 20, premin.features = 20, postmin.spots = 1, postmin.features = 10)

## User can retain the raw seuList by the following commond.  PRECASTObj <-
## CreatePRECASTObject(seuList, customGenelist=row.names(seuList[[1]]), rawData.preserve =
## TRUE)

Add the model setting

Add adjacency matrix list and parameter setting of PRECAST. More model setting parameters can be found in .

## check the number of genes/features after filtering step
PRECASTObj@seulist

## seuList is null since the default value `rawData.preserve` is FALSE.
PRECASTObj@seuList

## Add adjacency matrix list for a PRECASTObj object to prepare for PRECAST model fitting.
PRECASTObj <- AddAdjList(PRECASTObj, platform = "Visium")

## Add a model setting in advance for a PRECASTObj object. verbose =TRUE helps outputing the
## information in the algorithm.
PRECASTObj <- AddParSetting(PRECASTObj, Sigma_equal = FALSE, verbose = TRUE, maxIter = 30)

Fit PRECAST using this data

Fit PRECAST

For function PRECAST, users can specify the number of clusters K or set K to be an integer vector by using modified BIC(MBIC) to determine K. First, we try using user-specified number of clusters. For convenience, we give the selected number of clusters by MBIC (K=14).

### Given K
PRECASTObj <- PRECAST(PRECASTObj, K = 14)

Select a best model if K is an integer vector. Even if K is a scalar, this step is also neccessary to re-organize the results in PRECASTObj object.

## backup the fitting results in resList
resList <- PRECASTObj@resList
PRECASTObj <- SelectModel(PRECASTObj)

Integrate the two samples using the IntegrateSpaData function. For computational efficiency, this function exclusively integrates the variable genes. Specifically, in cases where users do not specify the PRECASTObj@seuList or seuList argument within the IntegrateSpaData function, it automatically focuses on integrating only the variable genes. The default setting for PRECASTObj@seuList is NULL when rawData.preserve in CreatePRECASTObject is set to FALSE. For instance:

print(PRECASTObj@seuList)
seuInt <- IntegrateSpaData(PRECASTObj, species = "Human")
seuInt
## The low-dimensional embeddings obtained by PRECAST are saved in PRECAST reduction slot.
Integrating all genes

There are two ways to use IntegrateSpaData integrating all genes, which will require more memory. We recommand running for all genes on server. The first one is to set value for PRECASTObj@seuList.

## assign the raw Seurat list object to it For illustration, we generate a new seuList with
## more genes; For integrating all genes, users can set `seuList <- bc2`.
genes <- c(row.names(PRECASTObj@seulist[[1]]), row.names(bc2[[1]])[1:10])
seuList <- lapply(bc2, function(x) x[genes, ])
PRECASTObj@seuList <- seuList  # 
seuInt <- IntegrateSpaData(PRECASTObj, species = "Human")
seuInt

The second method is to set a value for the argument seuList:

PRECASTObj@seuList <- NULL
## At the same time, we can set subsampling to speed up the computation.
seuInt <- IntegrateSpaData(PRECASTObj, species = "Human", seuList = seuList, subsample_rate = 0.5)
seuInt

First, user can choose a beautiful color schema using chooseColors().

cols_cluster <- chooseColors(palettes_name = "Classic 20", n_colors = 14, plot_colors = TRUE)

Show the spatial scatter plot for clusters


p12 <- SpaPlot(seuInt, item = "cluster", batch = NULL, point_size = 1, cols = cols_cluster, combine = TRUE,
    nrow.legend = 7)
p12
# users can plot each sample by setting combine=FALSE

Users can re-plot the above figures for specific need by returning a ggplot list object. For example, we plot the spatial heatmap using a common legend.

pList <- SpaPlot(seuInt, item = "cluster", batch = NULL, point_size = 1, cols = cols_cluster, combine = FALSE,
    nrow.legend = 7)
drawFigs(pList, layout.dim = c(1, 2), common.legend = TRUE, legend.position = "right", align = "hv")

Show the spatial UMAP/tNSE RGB plot to illustrate the performance in extracting features.

seuInt <- AddUMAP(seuInt)
p13 <- SpaPlot(seuInt, batch = NULL, item = "RGB_UMAP", point_size = 2, combine = TRUE, text_size = 15)
p13
# seuInt <- AddTSNE(seuInt) SpaPlot(seuInt, batch=NULL,item='RGB_TSNE',point_size=2,
# combine=T, text_size=15)

Show the tSNE plot based on the extracted features from PRECAST to check the performance of integration.

seuInt <- AddTSNE(seuInt, n_comp = 2)
p1 <- dimPlot(seuInt, item = "cluster", point_size = 0.5, font_family = "serif", cols = cols_cluster,
    border_col = "gray10", nrow.legend = 14, legend_pos = "right")  # Times New Roman
p2 <- dimPlot(seuInt, item = "batch", point_size = 0.5, font_family = "serif", legend_pos = "right")

drawFigs(list(p1, p2), layout.dim = c(1, 2), legend.position = "right", align = "hv")

Combined differential expression analysis

library(Seurat)
dat_deg <- FindAllMarkers(seuInt)
library(dplyr)
n <- 10
dat_deg %>%
    group_by(cluster) %>%
    top_n(n = n, wt = avg_log2FC) -> top10

seuInt <- ScaleData(seuInt)
seus <- subset(seuInt, downsample = 400)

Plot DE genes’ heatmap for each spatial domain identified by PRECAST.

color_id <- as.numeric(levels(Idents(seus)))
library(ggplot2)
## HeatMap
p1 <- doHeatmap(seus, features = top10$gene, cell_label = "Domain", grp_label = F, grp_color = cols_cluster[color_id],
    pt_size = 6, slot = "scale.data") + theme(legend.text = element_text(size = 10), legend.title = element_text(size = 13,
    face = "bold"), axis.text.y = element_text(size = 5, face = "italic", family = "serif"))
p1
Session Info 
sessionInfo()
#> R version 4.4.2 (2024-10-31)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.1 LTS
#> 
#> Matrix products: default
#> BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
#> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
#> 
#> locale:
#>  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
#>  [3] LC_TIME=en_US.UTF-8        LC_COLLATE=C              
#>  [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       
#> 
#> time zone: Etc/UTC
#> tzcode source: system (glibc)
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#> [1] rmarkdown_2.29
#> 
#> loaded via a namespace (and not attached):
#>  [1] digest_0.6.37     R6_2.5.1          fastmap_1.2.0     xfun_0.49        
#>  [5] maketools_1.3.1   cachem_1.1.0      knitr_1.49        htmltools_0.5.8.1
#>  [9] buildtools_1.0.0  lifecycle_1.0.4   cli_3.6.3         sass_0.4.9       
#> [13] jquerylib_0.1.4   compiler_4.4.2    sys_3.4.3         tools_4.4.2      
#> [17] evaluate_1.0.1    bslib_0.8.0       yaml_2.3.10       formatR_1.14     
#> [21] jsonlite_1.8.9    rlang_1.1.4