Motif analysis using topic modeling results for Cusanovich et al (2018) scATAC-seq data

Last updated: 2021-01-20

Checks: 7 0

Knit directory: scATACseq-topics/

This reproducible R Markdown analysis was created with workflowr (version 1.6.2). 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(20200729)

The command set.seed(20200729) 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: c9d6742

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 c9d6742. 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:    .Rhistory
    Ignored:    .Rproj.user/

Untracked files:
    Untracked:  analysis/diff_count_Buenrostro2018_chomVAR_scPeaks.Rmd
    Untracked:  analysis/gene_analysis_Cusanovich2018.Rmd
    Untracked:  analysis/process_data_Buenrostro2018_Chen2019.Rmd
    Untracked:  analysis/single_cell_rnaseq_demo.Rmd
    Untracked:  output/Cusanovich2018/
    Untracked:  output/clustering-Cusanovich2018.rds
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-TSS-sum.html
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-TSS-sum.png
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-TSS-sum_files/
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-genebody-sum.html
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-genebody-sum.png
    Untracked:  output/gsea_topic_1_Cusanovich2018-k=13-genebody-sum_files/
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-TSS-sum.html
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-TSS-sum.png
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-TSS-sum_files/
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-genebody-sum.html
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-genebody-sum.png
    Untracked:  output/volcano_topic_1_Cusanovich2018-k=13-genebody-sum_files/
    Untracked:  scripts/fit_all_models_Buenrostro_2018_chromVar_scPeaks_filtered.sbatch

Unstaged changes:
    Modified:   analysis/cisTopic_Buenrostro2018_chomVAR_scPeaks.Rmd
    Modified:   analysis/motif_gene_analysis_Cusanovich2018.Rmd
    Modified:   analysis/plots_Lareau2019_bonemarrow.Rmd
    Modified:   code/motif_analysis.R
    Modified:   code/plots.R
    Modified:   scripts/fit_all_models_Buenrostro_2018.sbatch
    Modified:   scripts/fit_cisTopic_Buenrostro_2018_chromVAR_scPeaks.sh
    Modified:   scripts/postfit_Buenrostro2018.sh

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/motif_analysis_Cusanovich2018.Rmd) and HTML (docs/motif_analysis_Cusanovich2018.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	c9d6742	kevinlkx	2021-01-20	updated the correlations between motif enrichment (using z-score) and gene z-score
html	1ab3d72	kevinlkx	2021-01-19	Build site.
Rmd	0134bdf	kevinlkx	2021-01-19	fix colors in scatter plots of create_motif_gene_scatterplot
html	a29293a	kevinlkx	2021-01-19	Build site.
Rmd	f51b56e	kevinlkx	2021-01-19	added colors to scatterplots for motif enrichment and gene scores
html	14eac34	kevinlkx	2021-01-19	Build site.
Rmd	5f18411	kevinlkx	2021-01-19	Plot motif enrichment and correlate with gene scores

Here we perform TF motif analysis for the Cusanovich et al (2018) scATAC-seq result inferred from the multinomial topic model with \(k = 13\).

Load packages and some functions used in this analysis

library(Matrix)
library(fastTopics)
library(dplyr)
library(tidyr)
library(ggplot2)
library(ggrepel)
library(cowplot)
library(plotly)
library(htmlwidgets)
library(DT)
library(reshape2)
source("code/motif_analysis.R")
source("code/plots.R")

Load data and topic model results

Load the data and the \(k = 13\) Poisson NMF fit results.

data.dir <- "/project2/mstephens/kevinluo/scATACseq-topics/data/Cusanovich_2018/processed_data/"
load(file.path(data.dir, "Cusanovich_2018.RData"))
rm(counts)

fit.dir <- "/project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018"
fit <- readRDS(file.path(fit.dir, "/fit-Cusanovich2018-scd-ex-k=13.rds"))$fit
fit_multinom <- poisson2multinom(fit)

Visualize by Structure plot grouped by tissues

set.seed(10)
colors_topics <- c("#a6cee3","#1f78b4","#b2df8a","#33a02c","#fb9a99","#e31a1c",
                   "#fdbf6f","#ff7f00","#cab2d6","#6a3d9a","#ffff99","#b15928",
                   "gray")
rows <- sample(nrow(fit$L),4000)
samples$tissue <- as.factor(samples$tissue)

p.structure <- structure_plot(select(fit_multinom,loadings = rows),
                                grouping = samples[rows, "tissue"],n = Inf,gap = 40,
                                perplexity = 50,topics = 1:13,colors = colors_topics,
                                num_threads = 4,verbose = FALSE)

print(p.structure)

Version	Author	Date
14eac34	kevinlkx	2021-01-19

Differential accessbility analysis of the ATAC-seq regions for the topics

Load results from differential accessbility analysis for the topics

out.dir <- "/project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018"
cat(sprintf("Load results from %s \n", out.dir))

# Load results from /project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018

diff_count_topics <- readRDS(file.path(out.dir, "/diffcount-Cusanovich2018-13topics.rds"))

Distribution of z-scores

zscore_topics <-  melt(diff_count_topics$Z)
colnames(zscore_topics) <- c("region", "topic", "zscore")
levels(zscore_topics$topic) <- colnames(diff_count_topics$Z)

z.quantile.99 <- apply(abs(diff_count_topics$Z), 2, quantile, 0.99)
cat("z-score 99% quantile: \n")
print(z.quantile.99)

p.hist.zscores <- ggplot(zscore_topics, aes(x=zscore)) + 
  geom_histogram(binwidth=1, color="black", fill="white") + 
  coord_cartesian(xlim = c(-10, 30)) + theme_cowplot(font_size = 10) +
  facet_wrap(~ topic, ncol=4)

print(p.hist.zscores)

Version	Author	Date
14eac34	kevinlkx	2021-01-19

# z-score 99% quantile: 
#       k1       k2       k3       k4       k5       k6       k7       k8 
# 21.42992 31.48751 25.46082 25.97670 34.64418 37.09098 32.07250 39.65746 
#       k9      k10      k11      k12      k13 
# 25.93102 15.88394 34.29782 39.80147 20.71928

Volcano plot of the regions

topic 1 examples

volcano_plot(diff_count_topics,k = 1,label_above_quantile = Inf, 
             subsample_below_quantile = 0.7, subsample_rate = 0.1)

# 161397 out of 436206 data points will be included in plot

Version	Author	Date
a29293a	kevinlkx	2021-01-19
14eac34	kevinlkx	2021-01-19

Motif enrichment analysis using HOMER

Details about HOMER motif analysis:
- http://homer.ucsd.edu/homer/motif/
- http://homer.ucsd.edu/homer/ngs/peakMotifs.html
Motif enrichment result using regions with z-score above 99% quantile.

Compile Homer results across topics

homer.dir <- paste0(out.dir, "/motifanalysis-Cusanovich2018-k=13-quantile/HOMER")
cat(sprintf("Directory of motif analysis result: %s \n", homer.dir))
homer_res_topics <- readRDS(file.path(homer.dir, "/homer_knownResults.rds"))
selected_regions <- readRDS(file.path(homer.dir, "/selected_regions.rds"))

# Compile Homer results (pvalue and ranking) across topics
motif_res <- compile_homer_motif_res(homer_res_topics)
saveRDS(motif_res, paste0(homer.dir, "/homer_motif_enrichment_results.rds"))
cat("compiled homer motif results are saved in", paste0(homer.dir, "/homer_motif_enrichment_results.rds"))

motif_table <- data.frame(motif = gsub("/.*", "", rownames(motif_res$mlog10P)),
                          motif_res$mlog10P)
DT::datatable(motif_table, rownames = F, caption = "Motif enrichment (-log10P)")

# Directory of motif analysis result: /project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018/motifanalysis-Cusanovich2018-k=13-quantile/HOMER 
# compiled homer motif results are saved in /project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018/motifanalysis-Cusanovich2018-k=13-quantile/HOMER/homer_motif_enrichment_results.rds

Heatmap of motif enrichment across topics

Clustering motifs by hierarchical clustering (motifs with similar enrichment across topics are plotted together)

create_motif_enrichment_heatmap(motif_res, cluster_motifs = TRUE, cluster_topics = FALSE, filter_motifs = TRUE, min_enrichment = 50, 
                                max_enrichment = 100, method_cluster = "average", font.size.motifs = 4, font.size.topics = 9)

Version	Author	Date
1ab3d72	kevinlkx	2021-01-19
a29293a	kevinlkx	2021-01-19
14eac34	kevinlkx	2021-01-19

# 114 out of 439 motifs included the heatmap

Cluster both motifs and topics by hierarchical clustering

create_motif_enrichment_heatmap(motif_res, cluster_motifs = TRUE, cluster_topics = TRUE, filter_motifs = TRUE, min_enrichment = 50, 
                                max_enrichment = 100, method_cluster = "average", font.size.motifs = 4, font.size.topics = 9)

Version	Author	Date
1ab3d72	kevinlkx	2021-01-19
a29293a	kevinlkx	2021-01-19
14eac34	kevinlkx	2021-01-19

# 114 out of 439 motifs included the heatmap

Scatterplots of motif enrichment

# Plot enrichment (-log10 p-value) and ranking of the motifs

plots <- vector("list", ncol(motif_res$mlog10P))
names(plots) <- colnames(motif_res$mlog10P)

for( i in 1:length(plots)){
 plots[[i]] <- create_motif_enrichment_ranking_plot(motif_res, k = i, 
                                                    max.overlaps = 15, subsample = FALSE)
}

# do.call(plot_grid,plots)

Plot motif enrichment (-log10 p-value) and the ranking

Topic 1 example

print(plots[[1]])

Version	Author	Date
1ab3d72	kevinlkx	2021-01-19
a29293a	kevinlkx	2021-01-19
14eac34	kevinlkx	2021-01-19

# Plot motif enrichment (-log10 p-value) in each topic vs other topics

plots <- vector("list", ncol(motif_res$mlog10P))
names(plots) <- colnames(motif_res$mlog10P)

for( i in 1:length(homer_res_topics)){
 plots[[i]] <- create_motif_enrichment_plot(motif_res, k = i, 
                                            max.overlaps = 15, subsample = TRUE)
}

# do.call(plot_grid,plots)

Plot motif enrichment (-log10 p-value) in topic 1 vs other topics

print(plots[[1]])

Version	Author	Date
1ab3d72	kevinlkx	2021-01-19
a29293a	kevinlkx	2021-01-19
14eac34	kevinlkx	2021-01-19

Motif enrichment vs gene score

Load pre-computed gene scores

gene.dir <- paste0(out.dir, "/geneanalysis-Cusanovich2018-k=13-TSS-l2")
cat(sprintf("Directory of gene analysis result: %s \n", gene.dir))
genescore_res <- readRDS(file.path(gene.dir, "genescore_result_topics.rds"))

genes <- genescore_res$genes
gene_scores <- genescore_res$Z
rownames(gene_scores) <- genes$SYMBOL
gene_logFC <- genescore_res$beta
rownames(gene_logFC) <- genes$SYMBOL

# Directory of gene analysis result: /project2/mstephens/kevinluo/scATACseq-topics/output/Cusanovich_2018/geneanalysis-Cusanovich2018-k=13-TSS-l2

Get TF genes

motif_names <- gsub("\\s*\\(.*", "", motif_res$motifs$motif)
gene_names <- genes$SYMBOL

TF_genes <- intersect(toupper(motif_names), toupper(gene_names))
cat(sprintf("%s TF genes mapped between motif names and gene symbol. \n", length(TF_genes)))

# 250 TF genes mapped between motif names and gene symbol.

Compute correlation between motif enrichment z-score and gene score:

Compute motif enrichment z-scores from the motif enrichment p-values
Plot motif enrichment (-log10 p-value) and correlation between motif enrichment z-score and gene scores
Rank motifs by motif enrichment (-log10 p-value) and correlation between motif enrichment z-score and gene scores