mist:methylation inference for single-cell along trajectory

Introduction

mist (Methylation Inference for Single-cell along Trajectory) is an R package for differential methylation (DM) analysis of single-cell DNA methylation (scDNAm) data. The package employs a Bayesian approach to model methylation changes along pseudotime and estimates developmental-stage-specific biological variations. It supports both single-group and two-group analyses, enabling users to identify genomic features exhibiting temporal changes in methylation levels or different methylation patterns between groups.

This vignette demonstrates how to use mist for: 1. Single-group analysis. 2. Two-group analysis.

Installation

To install the latest version of mist, run the following commands:

if (!requireNamespace("BiocManager", quietly = TRUE)) {
    install.packages("BiocManager")
}

# Install mist from GitHub
BiocManager::install("https://github.com/dxd429/mist")

From Bioconductor:

if (!requireNamespace("BiocManager", quietly = TRUE)) {
    install.packages("BiocManager")
}
BiocManager::install("mist")

To view the package vignette in HTML format, run the following lines in R:

library(mist)
vignette("mist")

Example Workflow for Single-Group Analysis

In this section, we will estimate parameters and perform differential methylation analysis using single-group data.

Step 1: Load Example Data

Here we load the example data from GSE121708.

library(mist)
library(SingleCellExperiment)
# Load sample scDNAm data
Dat_sce <- readRDS(system.file("extdata", "group1_sampleData_sce.rds", package = "mist"))

Step 2: Estimate Parameters Using estiParam

# Estimate parameters for single-group
Dat_sce <- estiParam(
    Dat_sce = Dat_sce,
    Dat_name = "Methy_level_group1",
    ptime_name = "pseudotime"
)

# Check the output
head(rowData(Dat_sce)$mist_pars)
##                      Beta_0      Beta_1      Beta_2      Beta_3       Beta_4
## ENSMUSG00000000001 1.247262 -0.46542216  0.40872069  0.27358090  0.016334836
## ENSMUSG00000000003 1.590342  1.64704281  2.69275466 -1.77234551 -2.915429081
## ENSMUSG00000000028 1.270999 -0.00676257  0.08666229  0.04560178  0.007623346
## ENSMUSG00000000037 1.013257 -4.27335989 11.42469890 -4.13755310 -3.003980014
## ENSMUSG00000000049 1.025324 -0.06262980  0.08890304  0.06549083  0.054392648
##                     Sigma2_1  Sigma2_2 Sigma2_3 Sigma2_4
## ENSMUSG00000000001  5.997288 15.413669 3.387755 1.879467
## ENSMUSG00000000003 25.630684  5.816801 5.600979 8.897011
## ENSMUSG00000000028  7.568958  7.204760 3.064988 2.409763
## ENSMUSG00000000037  8.266526 14.705174 6.751349 2.376986
## ENSMUSG00000000049  6.183266  8.779335 3.365732 1.135427

Step 3: Perform Differential Methylation Analysis Using dmSingle

# Perform differential methylation analysis for the single-group
Dat_sce <- dmSingle(Dat_sce)

# View the top genomic features with drastic methylation changes
head(rowData(Dat_sce)$mist_int)
## ENSMUSG00000000037 ENSMUSG00000000003 ENSMUSG00000000001 ENSMUSG00000000049 
##        0.060917767        0.032882889        0.011640344        0.006694779 
## ENSMUSG00000000028 
##        0.005477603

Step 4: Perform Differential Methylation Analysis Using plotGene

# Produce scatterplot with fitted curve of a specific gene
library(ggplot2)
plotGene(Dat_sce = Dat_sce,
         Dat_name = "Methy_level_group1",
         ptime_name = "pseudotime", 
         gene_name = "ENSMUSG00000000037")

Example Workflow for Two-Group Analysis

In this section, we will estimate parameters and perform DM analysis using data from two phenotypic groups.

Step 1: Load Two-Group Data

# Load two-group scDNAm data
Dat_sce_g1 <- readRDS(system.file("extdata", "group1_sampleData_sce.rds", package = "mist"))
Dat_sce_g2 <- readRDS(system.file("extdata", "group2_sampleData_sce.rds", package = "mist"))

Step 2: Estimate Parameters Using estiParam

# Estimate parameters for both groups
Dat_sce_g1 <- estiParam(
     Dat_sce = Dat_sce_g1,
     Dat_name = "Methy_level_group1",
     ptime_name = "pseudotime"
 )

Dat_sce_g2 <- estiParam(
     Dat_sce = Dat_sce_g2,
     Dat_name = "Methy_level_group2",
     ptime_name = "pseudotime"
 ) 

# Check the output
head(rowData(Dat_sce_g1)$mist_pars, n = 3)
##                      Beta_0      Beta_1     Beta_2      Beta_3        Beta_4
## ENSMUSG00000000001 1.244096 -0.73823702 0.59835487  0.40382076  0.0107327497
## ENSMUSG00000000003 1.579983  1.39525401 2.99723139 -1.61166547 -3.1187405973
## ENSMUSG00000000028 1.268161  0.01360779 0.05862375  0.02617795 -0.0006550722
##                     Sigma2_1  Sigma2_2 Sigma2_3 Sigma2_4
## ENSMUSG00000000001  5.987846 13.957408 3.926699 1.937718
## ENSMUSG00000000003 25.728956  4.376818 5.681909 9.027265
## ENSMUSG00000000028  7.748692  7.641173 3.190379 2.154878
head(rowData(Dat_sce_g2)$mist_pars, n = 3)
##                        Beta_0      Beta_1    Beta_2     Beta_3    Beta_4
## ENSMUSG00000000001  1.8934806 -0.37022945 3.3609942 -1.7153850 -1.406724
## ENSMUSG00000000003 -0.8142054 -0.71108294 2.1725988 -0.7481840 -0.655266
## ENSMUSG00000000028  2.3521512 -0.01854009 0.8514741 -0.1564616 -0.578869
##                     Sigma2_1  Sigma2_2 Sigma2_3 Sigma2_4
## ENSMUSG00000000001  5.882916  8.632136 3.696960 1.301908
## ENSMUSG00000000003  7.600704 10.449406 4.526966 2.893522
## ENSMUSG00000000028 11.655161  6.232757 3.960092 3.161971

Step 3: Perform Differential Methylation Analysis for Two-Group Comparison Using dmTwoGroups

# Perform DM analysis to compare the two groups
dm_results <- dmTwoGroups(
     Dat_sce_g1 = Dat_sce_g1,
     Dat_sce_g2 = Dat_sce_g2
 )

# View the top genomic features with different temporal patterns between groups
head(dm_results)
## ENSMUSG00000000037 ENSMUSG00000000003 ENSMUSG00000000001 ENSMUSG00000000049 
##        0.053337931        0.027937169        0.024367483        0.011251378 
## ENSMUSG00000000028 
##        0.003092907

Conclusion

mist provides a comprehensive suite of tools for analyzing scDNAm data along pseudotime, whether you are working with a single group or comparing two phenotypic groups. With the combination of Bayesian modeling and differential methylation analysis, mist is a powerful tool for identifying significant genomic features in scDNAm data.

Session info

## R version 4.5.2 (2025-10-31)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.3 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] stats4    stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
##  [1] ggplot2_4.0.0               SingleCellExperiment_1.32.0
##  [3] SummarizedExperiment_1.40.0 Biobase_2.70.0             
##  [5] GenomicRanges_1.62.0        Seqinfo_1.0.0              
##  [7] IRanges_2.44.0              S4Vectors_0.48.0           
##  [9] BiocGenerics_0.56.0         generics_0.1.4             
## [11] MatrixGenerics_1.22.0       matrixStats_1.5.0          
## [13] mist_1.2.0                  BiocStyle_2.38.0           
## 
## loaded via a namespace (and not attached):
##  [1] tidyselect_1.2.1         dplyr_1.1.4              farver_2.1.2            
##  [4] Biostrings_2.78.0        S7_0.2.0                 bitops_1.0-9            
##  [7] fastmap_1.2.0            RCurl_1.98-1.17          GenomicAlignments_1.46.0
## [10] XML_3.99-0.20            digest_0.6.37            lifecycle_1.0.4         
## [13] survival_3.8-3           magrittr_2.0.4           compiler_4.5.2          
## [16] rlang_1.1.6              sass_0.4.10              tools_4.5.2             
## [19] yaml_2.3.10              rtracklayer_1.70.0       knitr_1.50              
## [22] S4Arrays_1.10.0          labeling_0.4.3           curl_7.0.0              
## [25] DelayedArray_0.36.0      RColorBrewer_1.1-3       abind_1.4-8             
## [28] BiocParallel_1.44.0      withr_3.0.2              sys_3.4.3               
## [31] grid_4.5.2               scales_1.4.0             MASS_7.3-65             
## [34] mcmc_0.9-8               cli_3.6.5                mvtnorm_1.3-3           
## [37] rmarkdown_2.30           crayon_1.5.3             httr_1.4.7              
## [40] rjson_0.2.23             cachem_1.1.0             splines_4.5.2           
## [43] parallel_4.5.2           BiocManager_1.30.26      XVector_0.50.0          
## [46] restfulr_0.0.16          vctrs_0.6.5              Matrix_1.7-4            
## [49] jsonlite_2.0.0           SparseM_1.84-2           carData_3.0-5           
## [52] car_3.1-3                MCMCpack_1.7-1           Formula_1.2-5           
## [55] maketools_1.3.2          jquerylib_0.1.4          glue_1.8.0              
## [58] codetools_0.2-20         gtable_0.3.6             BiocIO_1.20.0           
## [61] tibble_3.3.0             pillar_1.11.1            htmltools_0.5.8.1       
## [64] quantreg_6.1             R6_2.6.1                 evaluate_1.0.5          
## [67] lattice_0.22-7           Rsamtools_2.26.0         cigarillo_1.0.0         
## [70] bslib_0.9.0              MatrixModels_0.5-4       coda_0.19-4.1           
## [73] SparseArray_1.10.1       xfun_0.54                buildtools_1.0.0        
## [76] pkgconfig_2.0.3