Enhancing droplet-based single-nucleus RNA-seq resolution using the semi-supervised machine learning classifier DIEM

Abstract

Single-nucleus RnA sequencing (snRnA-seq) measures gene expression in individual nuclei instead of cells, allowing for unbiased cell type characterization in solid tissues. We observe that snRNA-seq is commonly subject to contamination by high amounts of ambient RNA, which can lead to biased downstream analyses, such as identification of spurious cell types if overlooked. We present a novel approach to quantify contamination and filter droplets in snRNA-seq experiments, called Debris Identification using Expectation Maximization (DIEM). Our likelihood-based approach models the gene expression distribution of debris and cell types, which are estimated using EM. We evaluated DIEM using three snRNA-seq data sets: (1) human differentiating preadipocytes in vitro, (2) fresh mouse brain tissue, and (3) human frozen adipose tissue (AT) from six individuals. All three data sets showed evidence of extranuclear RNA contamination, and we observed that existing methods fail to account for contaminated droplets and led to spurious cell types. When compared to filtering using these state of the art methods, DIEM better removed droplets containing high levels of extranuclear RNA and led to higher quality clusters. Although DIEM was designed for snRNA-seq, our clustering strategy also successfully filtered single-cell RNA-seq data. To conclude, our novel method DIEM removes debris-contaminated droplets from single-cell-based data fast and effectively, leading to cleaner downstream analysis. Our code is freely available for use at https ://githu b.com/marca lva/ diem. Single-cell RNA sequencing (scRNA-seq) has grown considerably in use over the previous decade and permitted a transcriptomic view into the composition of heterogeneous mixtures of cells 1,2. Recent advances in droplet-based microfluidics have created a high-throughput opportunity to assay single cells by scaling up previous well-based technologies to tens to hundreds of thousands of cells 3. Single-nucleus RNA sequencing (snRNA-seq), where nuclei are used instead of cells, has allowed the critical extension of single-cell based technologies to solid tissues where isolation and suspension of individual cells is difficult or impossible 4. For example, snRNA-seq has been

Single-nuclei RNA sequencing uncovers non-cell autonomous changes in cerebellar astrocytes and oligodendrocytes that may contribute to Spinocerebellar Ataxia Type 1 (SCA1) pathogenesis

“…We identified a total of 47894 high-quality, snRNAseq profiles. Debris-contaminated nuclei were removed with Seurat using cutoff on number of genes per nuclei or mitochondrial gene percentage, as well as using semi-supervised machine learning classifier Debris Identification using Expectation Maximization (DIEM)[42]. We classified nuclei into the cell types with SingleR (v 1.6.1) using DropViz Cerebellum MetaCells reference [43] ( Figure 1B ).…”

Maternal-Fetal Immune Responses in Pregnant Women Infected with SARS-CoV-2

“…The fastq files were then aligned using kallisto 122 , and bustools 123 summarized the cell/gene transcript counts in a matrix for each sample using the “lamanno” workflow for scRNAseq. Each library was then processed using DIEM 124 to eliminate debris and empty droplets. In parallel, “cellranger counts” was also used to align the scRNAseq reads using the STAR 125 aligner to produce the bam files necessary for demultiplexing the individual of origin, based on the genotype information using souporcell 126 and demuxlet 127 .…”

Human liver single nucleus and single cell RNA sequencing identify a hepatocellular carcinoma-associated cell-type affecting survival

“…However, applying scRNA-seq technology to precious, archived human tissues, such as liver biopsies or resections, has proven to be challenging as it is not possible to dissociate intact cells from these existing biopsies of solid tissues. Single nucleus RNA sequencing (snRNA-seq) techniques [ 11 ] have overcome these technical challenges [ 12 ] and enabled cell-type level characterization of frozen solid tissues [ 13 – 16 ]. As scRNA-seq and snRNA-seq technologies improve, their use for solid tissues, such as liver, has expanded [ 17 , 18 ].…”