scReadSim: a single-cell RNA-seq and ATAC-seq read simulator

Abstract

Single-cell sequencing technologies emerged and diversified rapidly in the past few years, along with the successful development of many computational tools. Realistic simulators can help researchers benchmark computational tools. However, few simulators can generate single-cell multi-omics data, and none can generate reads directly. To fill in this gap, we propose scReadSim, a simulator for single-cell multi-omics reads. Trained on real data, scReadSim generates synthetic sequencing reads in BAM or FASTQ formats. We deployed scReadSim on a sci-ATAC-seq dataset and a single-cell multimodal dataset to show the resemblance between synthetic and real data at the read and count levels. Moreover, we show that scReadSim allows user-specified ground truths of accessible chromatin regions for single-cell chromatin accessibility data generation. In addition, scReadSim is flexible for allowing varying throughputs and library sizes as input parameters to guide experimental design.

A unified framework of realistic in silico data generation and statistical model inference for single-cell and spatial omics

“…S7). Moreover, coupled with our newly proposed read simulator scReadSim [26], scDesign3 extends the simulation of synthetic cells from the count level to the read level, unblocking its application for benchmarking read-level bioinformatics tools (Fig. 1g right).…”

“…For the 10x scATAC-seq dataset (ATAC), we used the R package Signac (version 1.7.0) [36] to first obtain a cell-by-peak matrix and then select 1133 differentially accessible peaks. For the sci-ATAC-seq data, the preprocessing and feature selection steps are described in [26]. For the 10x Visium dataset (VISIUM), we used the R package Seurat (version 4.1.1) to select the top 1000 spatially variable genes (SVGs).…”

“…Large r represent similar expression patterns in synthetic and test data. g , scDesign3 simulates a realistic sci-ATAC-seq dataset at both the count level (left: UMAP visualizations of real and synthetic cells in terms of peak counts) and the read level (right: pseudobulk read coverages; coupled with scReadSim [26]). h , scDesign3 generates a multi-omics (RNA expression + DNA methylation) dataset (right) by learning from real data that only have a single modality (left).…”

scReadSim: a single-cell RNA-seq and ATAC-seq read simulator

“…In the tutorials, we described the input and output formats, model parameters, and exemplary datasets for each functionality of scReadSim. The source code for reproducing the results are available at: https://github.com/Dominic7227/scReadSim_ source 49 .…”