Two step clustering analysis on LUHMES CROP-seq data

workflowr

• Summary
• Checks
• Past versions

Last updated: 2022-09-20

Checks: 7 0

Knit directory: GSFA_analysis/

This reproducible R Markdown analysis was created with workflowr (version 1.7.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

---

R Markdown file: up-to-date

Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

Environment: empty

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

Seed: set.seed(20220524)

The command set.seed(20220524) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Session information: recorded

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Cache: none

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

File paths: relative

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Repository version: 9cc77c5

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 9cc77c5. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:

Ignored files:
    Ignored:    .DS_Store
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/

Untracked files:
    Untracked:  Rplots.pdf
    Untracked:  analysis/check_Tcells_datasets.Rmd
    Untracked:  analysis/fscLVM_analysis.Rmd
    Untracked:  analysis/spca_LUHMES_data.Rmd
    Untracked:  analysis/test_seurat.Rmd
    Untracked:  code/gsfa_negctrl_job.sbatch
    Untracked:  code/music_LUHMES_Yifan.R
    Untracked:  code/plotting_functions.R
    Untracked:  code/run_fscLVM_LUHMES_data.R
    Untracked:  code/run_gsfa_2groups_negctrl.R
    Untracked:  code/run_gsfa_negctrl.R
    Untracked:  code/run_music_LUHMES.R
    Untracked:  code/run_music_LUHMES_data.sbatch
    Untracked:  code/run_music_LUHMES_data_20topics.R
    Untracked:  code/run_music_LUHMES_data_20topics.sbatch
    Untracked:  code/run_sceptre_Tcells_data.sbatch
    Untracked:  code/run_sceptre_Tcells_stimulated_data.sbatch
    Untracked:  code/run_sceptre_Tcells_test_data.sbatch
    Untracked:  code/run_sceptre_Tcells_unstimulated_data.sbatch
    Untracked:  code/run_sceptre_permuted_data.sbatch
    Untracked:  code/run_spca_LUHMES.R
    Untracked:  code/run_spca_TCells.R
    Untracked:  code/run_twostep_clustering_LUHMES_data.sbatch
    Untracked:  code/run_twostep_clustering_Tcells_data.sbatch
    Untracked:  code/run_unguided_gsfa_LUHMES.R
    Untracked:  code/run_unguided_gsfa_LUHMES.sbatch
    Untracked:  code/run_unguided_gsfa_Tcells.R
    Untracked:  code/run_unguided_gsfa_Tcells.sbatch
    Untracked:  code/sceptre_LUHMES_data.R
    Untracked:  code/sceptre_Tcells_stimulated_data.R
    Untracked:  code/sceptre_Tcells_unstimulated_data.R
    Untracked:  code/sceptre_permutation_analysis.R
    Untracked:  code/sceptre_permute_analysis.R
    Untracked:  code/seurat_sim_fpr_tpr.R
    Untracked:  code/unguided_GFSA_mixture_normal_prior.cpp

Unstaged changes:
    Modified:   analysis/sceptre_TCells_data.Rmd
    Modified:   code/run_sceptre_LUHMES_data.R
    Modified:   code/run_sceptre_LUHMES_data.sbatch
    Modified:   code/run_sceptre_LUHMES_permuted_data.R
    Modified:   code/run_sceptre_Tcells_permuted_data.R
    Modified:   code/run_sceptre_cropseq_data.sbatch
    Modified:   code/run_twostep_clustering_LUHMES_data.R
    Modified:   code/sceptre_analysis.R

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

---

These are the previous versions of the repository in which changes were made to the R Markdown (analysis/twostep_clustering_LUHMES_data.Rmd) and HTML (docs/twostep_clustering_LUHMES_data.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File	Version	Author	Date	Message
Rmd	9cc77c5	kevinlkx	2022-09-20	updated the size of qq plots
html	d8ffeac	kevinlkx	2022-09-20	Build site.
html	a112e26	kevinlkx	2022-09-20	Build site.
Rmd	d432d4a	kevinlkx	2022-09-20	updated the size of qq plots
html	1dc83b4	kevinlkx	2022-09-08	Build site.
Rmd	38ebf7d	kevinlkx	2022-09-08	updated the Seurat results
html	deb23d1	kevinlkx	2022-08-31	Build site.
Rmd	ad6135c	kevinlkx	2022-08-31	added qqplots for combined results
html	040c06a	kevinlkx	2022-08-25	Build site.
Rmd	9894017	kevinlkx	2022-08-25	added qqplots
html	14111b7	kevinlkx	2022-08-12	Build site.
Rmd	71a5d69	kevinlkx	2022-08-12	added more descriptions
html	12f1a27	kevinlkx	2022-08-12	Build site.
Rmd	9212bba	kevinlkx	2022-08-12	added more descriptions
html	d7c7e29	kevinlkx	2022-08-12	Build site.
Rmd	0bc181b	kevinlkx	2022-08-12	added more descriptions
html	056e9c8	kevinlkx	2022-08-12	Build site.
Rmd	a900cfd	kevinlkx	2022-08-12	added more descriptions
html	02e29d2	kevinlkx	2022-08-11	Build site.
Rmd	5796838	kevinlkx	2022-08-11	fix intro formating
html	da6979c	kevinlkx	2022-08-11	Build site.
Rmd	39baeef	kevinlkx	2022-08-11	update barplots for DEGs

---

mkdir -p /project2/xinhe/kevinluo/GSFA/data

cp /project2/xinhe/yifan/Factor_analysis/shared_data/LUHMES_cropseq_data_seurat.rds \
  /project2/xinhe/kevinluo/GSFA/data

cp /project2/xinhe/yifan/Factor_analysis/LUHMES/GSE142078_raw/GSM4219575_Run1_genes.tsv.gz \
  /project2/xinhe/kevinluo/GSFA/data/LUHMES_GSM4219575_Run1_genes.tsv.gz

Scripts for running the analysis:

• R script.

mkdir -p /project2/xinhe/kevinluo/GSFA/twostep_clustering/log
cd /project2/xinhe/kevinluo/GSFA/twostep_clustering/log

sbatch --mem=40G ~/projects/GSFA_analysis/code/run_twostep_clustering_LUHMES_data.sbatch

Load the data sets

CROP-seq datasets: /project2/xinhe/yifan/Factor_analysis/shared_data/LUHMES_cropseq_data_seurat.rds

The data are Seurat objects, with raw gene counts stored in obj@assays$RNA@counts, and cell meta data stored in obj@meta.data. Normalized and scaled data used for GSFA are stored in obj@assays$RNA@scale.data, the rownames of which are the 6k genes used for GSFA.

Load packages

suppressPackageStartupMessages(library(data.table))
dyn.load('/software/geos-3.7.0-el7-x86_64/lib64/libgeos_c.so') # attach the geos lib for Seurat
suppressPackageStartupMessages(library(Seurat))
suppressPackageStartupMessages(library(ComplexHeatmap))
suppressPackageStartupMessages(library(ggplot2))
require(reshape2)
require(dplyr)
require(ComplexHeatmap)
theme_set(theme_bw() + theme(plot.title = element_text(size = 14, hjust = 0.5),
                             axis.title = element_text(size = 14),
                             axis.text = element_text(size = 13),
                             legend.title = element_text(size = 13),
                             legend.text = element_text(size = 12),
                             panel.grid.minor = element_blank())
)
suppressPackageStartupMessages(library(gridExtra))
source("code/plotting_functions.R")

Set directories

data_dir <- "/project2/xinhe/kevinluo/GSFA/data/"
res_dir <- "/project2/xinhe/kevinluo/GSFA/twostep_clustering/LUHMES/"
dir.create(res_dir, recursive = TRUE, showWarnings = FALSE)

Load two-step clustering analysis result

combined_obj <- readRDS(file.path(res_dir, "LUHMES_seurat_clustered.rds"))

Associate perturbations with clusters

perturb_matrix <- combined_obj@meta.data[, 4:18]

cluster_labels <- Idents(combined_obj)
cluster_labels <- as.factor(as.numeric(as.character(cluster_labels))+1)
new_cluster_labels <- paste0("k", levels(cluster_labels))
names(new_cluster_labels) <- levels(combined_obj)
combined_obj <- RenameIdents(combined_obj, new_cluster_labels)

cluster_matrix <- matrix(0, nrow = nrow(perturb_matrix), ncol = length(levels(cluster_labels)))
cluster_matrix[cbind(1:nrow(perturb_matrix), cluster_labels)] <- 1
rownames(cluster_matrix) <- rownames(perturb_matrix)
colnames(cluster_matrix) <- new_cluster_labels

Use Chi-squared tests for the association of perturbations and clusters (2 x 2 tables)

summary_df <- expand.grid(colnames(perturb_matrix), colnames(cluster_matrix))
colnames(summary_df) <- c("perturb", "cluster")

summary_df <- cbind(summary_df, statistic = NA, stdres = NA, pval = NA)

for(i in 1:nrow(summary_df)){
  dt <- table(data.frame(perturb = perturb_matrix[,summary_df$perturb[i]], 
                         cluster = cluster_matrix[,summary_df$cluster[i]]))
  chisq <- chisq.test(dt)
  summary_df[i, ]$statistic <- chisq$statistic
  summary_df[i, ]$stdres <- chisq$stdres[2,2]
  summary_df[i, ]$pval <- chisq$p.value
}

summary_df$fdr <- p.adjust(summary_df$pval, method = "BH")
summary_df$bonferroni_adj <- p.adjust(summary_df$pval, method = "bonferroni")

# saveRDS(summary_df, file = file.path(res_dir, "LUHMES_seurat_guide_cluster_chisq_summary_df.rds"))

stdres_mat <- reshape2::dcast(summary_df %>% dplyr::select(perturb, cluster, stdres), perturb ~ cluster, value.var = "stdres")
rownames(stdres_mat) <- stdres_mat$perturb
stdres_mat$perturb <- NULL

fdr_mat <- reshape2::dcast(summary_df %>% dplyr::select(perturb, cluster, fdr), perturb ~ cluster, value.var = "fdr")
rownames(fdr_mat) <- fdr_mat$perturb
fdr_mat$perturb <- NULL

bonferroni_mat <- reshape2::dcast(summary_df %>% dplyr::select(perturb, cluster, bonferroni_adj), 
                                  perturb ~ cluster, value.var = "bonferroni_adj")
rownames(bonferroni_mat) <- bonferroni_mat$perturb
bonferroni_mat$perturb <- NULL

KO_names <- rownames(fdr_mat)
dotplot_effectsize(t(stdres_mat), t(fdr_mat),
                   reorder_markers = c(KO_names[KO_names!="Nontargeting"], "Nontargeting"),
                   color_lgd_title = "Chi-squared test\nstandardized residuals",
                   size_lgd_title = "FDR",
                   max_score = 4,
                   min_score = -4,
                   by_score = 2) + coord_flip()

Past versions of stdres-fdr-dotplot-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
da6979c	kevinlkx	2022-08-11

Plot perturbation ~ cluster associations (show Bonferroni adjusted p-values)

KO_names <- rownames(bonferroni_mat)
dotplot_effectsize(t(stdres_mat), t(bonferroni_mat),
                   reorder_markers = c(KO_names[KO_names!="Nontargeting"], "Nontargeting"),
                   color_lgd_title = "Chi-squared test\nstandardized residuals",
                   size_lgd_title = "Bonferroni\nadjusted p-value",
                   max_score = 4,
                   min_score = -4,
                   by_score = 2) + coord_flip()

Past versions of stdres-bonferroni-dotplot-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
da6979c	kevinlkx	2022-08-11

Find DE genes for each cluster and assign DE genes to associated perturbations

First, find DE genes for each cluster using MAST (Bonferroni adjusted p-values < 0.05), Then, for each perturbation, find the associated clusters, and pull the DE genes for those clusters.

feature.names <- data.frame(fread(file.path(data_dir, "LUHMES_GSM4219575_Run1_genes.tsv.gz"),
                                  header = FALSE), stringsAsFactors = FALSE)

de.markers <- readRDS(file.path(res_dir, "LUHMES_seurat_MAST_DEGs.rds"))
names(de.markers) <- paste0("k", levels(cluster_labels))

de.genes.clusters <- vector("list", length = length(de.markers))
names(de.genes.clusters) <- names(de.markers)
for( i in 1:length(de.genes.clusters)){
  de_sumstats <- de.markers[[i]]
  de_genes <- unique(rownames(de_sumstats[de_sumstats$p_val_adj < 0.05,]))
  # de_genes <- feature.names$V2[match(de_genes, feature.names$V1)]
  de.genes.clusters[[i]] <- de_genes
}

Number of DE genes for each perturbation (Chi-squared test FDR < 0.05)

perturb_names <- colnames(perturb_matrix)
perturb_names <- c("Nontargeting", perturb_names[perturb_names!="Nontargeting"])

de.genes.perturbs <- vector("list", length = length(perturb_names))
names(de.genes.perturbs) <- perturb_names

for(i in 1:length(de.genes.perturbs)){
  perturb_name <- names(de.genes.perturbs)[i]
  associated_cluster_labels <- colnames(fdr_mat)[which(fdr_mat[perturb_name, ] < 0.05)]
  if(length(associated_cluster_labels) > 0){
    de.genes.perturbs[[i]] <- unique(unlist(de.genes.clusters[associated_cluster_labels]))
  }
}

num.de.genes.perturbs <- sapply(de.genes.perturbs, length)

twostep_clustering_fdr0.05_genes <- de.genes.perturbs

dge_plot_df <- data.frame(Perturbation = names(num.de.genes.perturbs), Num_DEGs = num.de.genes.perturbs)
dge_plot_df$Perturbation <- factor(dge_plot_df$Perturbation, levels = names(num.de.genes.perturbs))

ggplot(data=dge_plot_df, aes(x = Perturbation, y = Num_DEGs+1)) +
  geom_bar(position = "dodge", stat = "identity") +
  geom_text(aes(label = Num_DEGs), position=position_dodge(width=0.9), vjust=-0.25) +
  scale_y_log10() +
  scale_fill_brewer(palette = "Set2") +
  labs(x = "Target gene",
       y = "Number of DEGs",
       title = "Number of DEGs detected by Two-step clustering with MAST DE analysis") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1, size = 12),
        legend.position = "bottom",
        legend.text = element_text(size = 13))

Past versions of count-de-genes-fdr-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
da6979c	kevinlkx	2022-08-11

Number of DE genes for each perturbation (Chi-squared test Bonferroni adjusted p-value < 0.05)

perturb_names <- colnames(perturb_matrix)
perturb_names <- c("Nontargeting", perturb_names[perturb_names!="Nontargeting"])

de.genes.perturbs <- vector("list", length = length(perturb_names))
names(de.genes.perturbs) <- perturb_names

for(i in 1:length(de.genes.perturbs)){
  perturb_name <- names(de.genes.perturbs)[i]
  associated_cluster_labels <- colnames(bonferroni_mat)[which(bonferroni_mat[perturb_name, ] < 0.05)]
  if(length(associated_cluster_labels) > 0){
    de.genes.perturbs[[i]] <- unique(unlist(de.genes.clusters[associated_cluster_labels]))
  }
}

num.de.genes.perturbs <- sapply(de.genes.perturbs, length)

twostep_clustering_bonferroni0.05_genes <- de.genes.perturbs

dge_plot_df <- data.frame(Perturbation = names(num.de.genes.perturbs), Num_DEGs = num.de.genes.perturbs)
dge_plot_df$Perturbation <- factor(dge_plot_df$Perturbation, levels = names(num.de.genes.perturbs))

ggplot(data=dge_plot_df, aes(x = Perturbation, y = Num_DEGs+1)) +
  geom_bar(position = "dodge", stat = "identity") +
  geom_text(aes(label = Num_DEGs), position=position_dodge(width=0.9), vjust=-0.25) +
  scale_y_log10() +
  scale_fill_brewer(palette = "Set2") +
  labs(x = "Target gene",
       y = "Number of DEGs",
       title = "Number of DEGs detected by Two-step clustering with MAST DE analysis") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1, size = 12),
        legend.position = "bottom",
        legend.text = element_text(size = 13))

Past versions of count-de-genes-bonferroni-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
da6979c	kevinlkx	2022-08-11

Compare single-gene DE p-value distributions between two-step clustering analysis and GSFA

fdr_cutoff <- 0.05
lfsr_cutoff <- 0.05

Load the output of GSFA fit_gsfa_multivar() run.

data_folder <- "/project2/xinhe/yifan/Factor_analysis/LUHMES/"
fit <- readRDS(paste0(data_folder,
                      "gsfa_output_detect_01/use_negctrl/All.gibbs_obj_k20.svd_negctrl.seed_14314.light.rds"))
gibbs_PM <- fit$posterior_means
lfsr_mat <- fit$lfsr[, -ncol(fit$lfsr)]
total_effect <- fit$total_effect[, -ncol(fit$total_effect)]
KO_names <- colnames(lfsr_mat)

DEGs detected by GSFA

        ADNP       ARID1B        ASH1L         CHD2         CHD8       CTNND2 
         795          310          322          756            0            0 
      DYRK1A        HDAC5        MECP2        MYT1L Nontargeting         POGZ 
          23            0            0            0            0            0 
        PTEN         RELN        SETD5 
         895            0          466 

Load MAST single-gene DE result

guides <- KO_names[KO_names!="Nontargeting"]

mast_list <- list()
for (m in guides){
  fname <- paste0(data_folder, "processed_data/MAST/dev_top6k_negctrl/gRNA_", m, ".dev_res_top6k.vs_negctrl.rds")
  tmp_df <- readRDS(fname)
  tmp_df$geneID <- rownames(tmp_df)
  tmp_df <- tmp_df %>% dplyr::rename(FDR = fdr, PValue = pval)
  mast_list[[m]] <- tmp_df
}
mast_signif_counts <- sapply(mast_list, function(x){filter(x, FDR < fdr_cutoff) %>% nrow()})

QQ-plots of MAST DE p-values for the GSFA significant genes vs two-step clustering DE genes.

qqplots <- list()
for(i in 1:length(guides)){
  guide <- guides[i]
  mast_res <- mast_list[[guide]]
  tsc_de_genes <- twostep_clustering_fdr0.05_genes[[guide]]
  gsfa_de_genes <- gsfa_sig_genes[[guide]]
  tsc_de_genes <- intersect(tsc_de_genes, rownames(mast_res))
  gsfa_de_genes <- intersect(gsfa_de_genes, rownames(mast_res))
  
  if(length(tsc_de_genes)>0 && length(gsfa_de_genes) >0){
    mast_res$tsc_gene <- 0
    mast_res[tsc_de_genes, ]$tsc_gene <- 1
    mast_res$gsfa_gene <- 0
    mast_res[gsfa_de_genes, ]$gsfa_gene <- 1
    pvalue_list <- list('Two-step clustering'=dplyr::filter(mast_res,tsc_gene==1)$PValue,
                        'GSFA'=dplyr::filter(mast_res,gsfa_gene==1)$PValue,
                        'all genes'=mast_res$PValue)
    qqplots[[guide]] <- qqplot.pvalue(pvalue_list, pointSize = 1, legendSize = 4) + 
      ggtitle(guide) + theme(plot.title = element_text(hjust = 0.5)) +
      scale_colour_discrete(name="Method")
  }
}

grid.arrange(grobs = qqplots, nrow = 2, ncol = 2)

Past versions of qqplots-gsfa-tsc-all-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
deb23d1	kevinlkx	2022-08-31
040c06a	kevinlkx	2022-08-25

Pooling p-values from all perturbations

combined_mast_res <- data.frame()
for(i in 1:length(guides)){
  guide <- guides[i]
  mast_res <- mast_list[[guide]]
  tsc_de_genes <- twostep_clustering_fdr0.05_genes[[guide]]
  gsfa_de_genes <- gsfa_sig_genes[[guide]]
  tsc_de_genes <- intersect(tsc_de_genes, rownames(mast_res))
  gsfa_de_genes <- intersect(gsfa_de_genes, rownames(mast_res))
  mast_res$tsc_gene <- 0
  if(length(tsc_de_genes) >0){
    mast_res[tsc_de_genes, ]$tsc_gene <- 1
  }
  mast_res$gsfa_gene <- 0
  if(length(gsfa_de_genes) >0){
    mast_res[gsfa_de_genes, ]$gsfa_gene <- 1
  }
  combined_mast_res <- rbind(combined_mast_res, mast_res)
}

pvalue_list <- list('Two-step clustering'=dplyr::filter(combined_mast_res,tsc_gene==1)$PValue,
                    'GSFA'=dplyr::filter(combined_mast_res,gsfa_gene==1)$PValue,
                    'all genes'=combined_mast_res$PValue)

qqplot.pvalue(pvalue_list, pointSize = 1, legendSize = 4) + 
      ggtitle("") + theme(plot.title = element_text(hjust = 0.5)) +
      scale_colour_discrete(name="Method")

Past versions of qqplots-gsfa-tsc-all-combined-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
deb23d1	kevinlkx	2022-08-31

QQ-plots of MAST DE p-values for the GSFA only genes vs two-step only genes.

qqplots <- list()
for(i in 1:length(guides)){
  guide <- guides[i]
  mast_res <- mast_list[[guide]]
  tsc_de_genes <- twostep_clustering_fdr0.05_genes[[guide]]
  gsfa_de_genes <- gsfa_sig_genes[[guide]]
  tsc_de_genes <- intersect(tsc_de_genes, rownames(mast_res))
  gsfa_de_genes <- intersect(gsfa_de_genes, rownames(mast_res))
  
  if(length(tsc_de_genes)>0 && length(gsfa_de_genes) >0){
    mast_res$tsc_only_gene <- 0
    mast_res[setdiff(tsc_de_genes, gsfa_de_genes), ]$tsc_only_gene <- 1
    mast_res$gsfa_only_gene <- 0
    mast_res[setdiff(gsfa_de_genes, tsc_de_genes), ]$gsfa_only_gene <- 1
    
    pvalue_list <- list('Two-step clustering only'=dplyr::filter(mast_res,tsc_only_gene==1)$PValue,
                        'GSFA only'=dplyr::filter(mast_res,gsfa_only_gene==1)$PValue,
                        'all genes'=mast_res$PValue)
    qqplots[[guide]] <- qqplot.pvalue(pvalue_list, pointSize = 1, legendSize = 4) + 
      ggtitle(guide) + theme(plot.title = element_text(hjust = 0.5)) +
      scale_colour_discrete(name="Method")

  }
}

grid.arrange(grobs = qqplots, nrow = 2, ncol = 2)

Past versions of qqplots-gsfa-tsc-only-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
deb23d1	kevinlkx	2022-08-31
040c06a	kevinlkx	2022-08-25

Pooling p-values from all perturbations

combined_mast_res <- data.frame()
for(i in 1:length(guides)){
  guide <- guides[i]
  mast_res <- mast_list[[guide]]
  tsc_de_genes <- twostep_clustering_fdr0.05_genes[[guide]]
  gsfa_de_genes <- gsfa_sig_genes[[guide]]
  tsc_de_genes <- intersect(tsc_de_genes, rownames(mast_res))
  gsfa_de_genes <- intersect(gsfa_de_genes, rownames(mast_res))
  mast_res$tsc_only_gene <- 0
  if(length(setdiff(tsc_de_genes, gsfa_de_genes)) >0){
    mast_res[setdiff(tsc_de_genes, gsfa_de_genes), ]$tsc_only_gene <- 1
  }
  mast_res$gsfa_only_gene <- 0
  if(length(setdiff(gsfa_de_genes, tsc_de_genes)) >0){
    mast_res[setdiff(gsfa_de_genes, tsc_de_genes), ]$gsfa_only_gene <- 1
  }
  combined_mast_res <- rbind(combined_mast_res, mast_res)
}

pvalue_list <- list('Two-step clustering only'=dplyr::filter(combined_mast_res,tsc_only_gene==1)$PValue,
                    'GSFA only'=dplyr::filter(combined_mast_res,gsfa_only_gene==1)$PValue,
                    'all genes'=combined_mast_res$PValue)

qqplot.pvalue(pvalue_list, pointSize = 1, legendSize = 4) + 
      ggtitle("") + theme(plot.title = element_text(hjust = 0.5)) +
      scale_colour_discrete(name="Method")

Past versions of qqplots-gsfa-tsc-only-combined-1.png

Version	Author	Date
1dc83b4	kevinlkx	2022-09-08
deb23d1	kevinlkx	2022-08-31

Session information

sessionInfo()

R version 4.2.0 (2022-04-22)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS/LAPACK: /software/openblas-0.3.13-el7-x86_64/lib/libopenblas_haswellp-r0.3.13.so

locale:
 [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C         LC_TIME=C           
 [4] LC_COLLATE=C         LC_MONETARY=C        LC_MESSAGES=C       
 [7] LC_PAPER=C           LC_NAME=C            LC_ADDRESS=C        
[10] LC_TELEPHONE=C       LC_MEASUREMENT=C     LC_IDENTIFICATION=C 

attached base packages:
[1] grid      stats     graphics  grDevices utils     datasets  methods  
[8] base     

other attached packages:
 [1] lattice_0.20-45       gridExtra_2.3         dplyr_1.0.9          
 [4] reshape2_1.4.4        ggplot2_3.3.6         ComplexHeatmap_2.12.0
 [7] sp_1.4-7              SeuratObject_4.1.0    Seurat_4.1.1         
[10] data.table_1.14.2     workflowr_1.7.0      

loaded via a namespace (and not attached):
  [1] circlize_0.4.15       plyr_1.8.7            igraph_1.3.4         
  [4] lazyeval_0.2.2        splines_4.2.0         listenv_0.8.0        
  [7] scattermore_0.8       digest_0.6.29         foreach_1.5.2        
 [10] htmltools_0.5.2       fansi_1.0.3           magrittr_2.0.3       
 [13] tensor_1.5            cluster_2.1.3         doParallel_1.0.17    
 [16] ROCR_1.0-11           globals_0.15.0        matrixStats_0.62.0   
 [19] R.utils_2.11.0        spatstat.sparse_2.1-1 colorspace_2.0-3     
 [22] ggrepel_0.9.1         xfun_0.30             callr_3.7.0          
 [25] crayon_1.5.1          jsonlite_1.8.0        progressr_0.10.0     
 [28] spatstat.data_2.2-0   survival_3.3-1        zoo_1.8-10           
 [31] iterators_1.0.14      glue_1.6.2            polyclip_1.10-0      
 [34] gtable_0.3.0          leiden_0.4.2          GetoptLong_1.0.5     
 [37] shape_1.4.6           future.apply_1.9.0    BiocGenerics_0.42.0  
 [40] abind_1.4-5           scales_1.2.0          DBI_1.1.3            
 [43] spatstat.random_2.2-0 miniUI_0.1.1.1        Rcpp_1.0.8.3         
 [46] viridisLite_0.4.0     xtable_1.8-4          clue_0.3-61          
 [49] reticulate_1.24       spatstat.core_2.4-2   stats4_4.2.0         
 [52] htmlwidgets_1.5.4     httr_1.4.3            RColorBrewer_1.1-3   
 [55] ellipsis_0.3.2        ica_1.0-2             R.methodsS3_1.8.1    
 [58] farver_2.1.0          pkgconfig_2.0.3       sass_0.4.1           
 [61] uwot_0.1.11           deldir_1.0-6          utf8_1.2.2           
 [64] labeling_0.4.2        tidyselect_1.1.2      rlang_1.0.2          
 [67] later_1.3.0           munsell_0.5.0         tools_4.2.0          
 [70] cli_3.3.0             generics_0.1.2        ggridges_0.5.3       
 [73] evaluate_0.15         stringr_1.4.0         fastmap_1.1.0        
 [76] yaml_2.3.5            goftest_1.2-3         processx_3.5.3       
 [79] knitr_1.39            fs_1.5.2              fitdistrplus_1.1-8   
 [82] purrr_0.3.4           RANN_2.6.1            pbapply_1.5-0        
 [85] future_1.25.0         nlme_3.1-157          whisker_0.4          
 [88] mime_0.12             R.oo_1.24.0           compiler_4.2.0       
 [91] rstudioapi_0.13       plotly_4.10.0         png_0.1-7            
 [94] spatstat.utils_2.3-1  tibble_3.1.7          bslib_0.3.1          
 [97] stringi_1.7.6         highr_0.9             ps_1.7.0             
[100] rgeos_0.5-9           Matrix_1.4-1          vctrs_0.4.1          
[103] pillar_1.7.0          lifecycle_1.0.1       spatstat.geom_2.4-0  
[106] lmtest_0.9-40         jquerylib_0.1.4       GlobalOptions_0.1.2  
[109] RcppAnnoy_0.0.19      cowplot_1.1.1         irlba_2.3.5          
[112] httpuv_1.6.5          patchwork_1.1.1       R6_2.5.1             
[115] promises_1.2.0.1      KernSmooth_2.23-20    IRanges_2.30.0       
[118] parallelly_1.31.1     codetools_0.2-18      MASS_7.3-56          
[121] assertthat_0.2.1      rprojroot_2.0.3       rjson_0.2.21         
[124] withr_2.5.0           sctransform_0.3.3     S4Vectors_0.34.0     
[127] mgcv_1.8-40           parallel_4.2.0        rpart_4.1.16         
[130] tidyr_1.2.0           rmarkdown_2.14        Rtsne_0.16           
[133] git2r_0.30.1          getPass_0.2-2         shiny_1.7.1