computationally investigates both the DNA methylation and copy
number states from scRRBS sequencing libraries and in parallel
measures cytoplasmic transcript levels.
In summary, the integration of genomic, epigenomic, tran-
scriptomic data are emerging as realistic prospect, benefiting from
rapidly developing sequencing technology. For example, Pacific
Biosciences and nanopore sequencers can interpret more than one
analyte in parallel [77, 78]. Thus, we expect that future sequencing
and analyzing approaches may provide the crucial advances that
expand the refine single-cell multiomics to the insights into gen-
erating the comprehensive atlases of cell states and lineages for
cellular systems, ranging from tissue microenvironment to whole
organisms.2 Computational Methods and Challenges for Single-Cell Transcriptomics
The development of high-throughput scRNA-seq has already led to
profound insights into biology, ranging from the identification of
novel cell types or subclones to global patterns of stochastic gene
expression for cell-to-cell heterogeneity or reconstructing cell dif-
ferentiation trajectories. Alongside the technological break-
throughs that have facilitated the large-scale generation of single-
cell transcriptomic data, it is crucial to apply appropriate computa-
tional and/or statistical methods to ensure that scRNA-seq data are
fully exploited and interpreted correctly. Although some tools for
analyzing RNA-seq data from bulk samples can be readily applied to
scRNA-seq data, many new computational strategies are required
to fully exploit this data type and to enable a comprehensive yet
detailed study of gene expression at the single-cell level. Generally,
as the same as the bulk RNA-seq data, three types of steps are
implemented to interrogate scRNA-seq data (Fig.3): alignment
and quality control (QC), normalization and quantification, and
applications for biological insights. The first steps (orange) are
general for any high-throughput sequencing data. Later steps
(blue) require a mix of existing RNA-seq analysis methods and
novel methods to address the technical difference of scRNA-seq.
The biological interpretation (red) should be analyzed with meth-
ods specifically developed for scRNA-seq. In this section, we survey
the various computational methods that are applied specifically to
the multiple steps of scRNA-seq data analysis aiming at deciphering
the transcriptomic dynamics at single-cell level (Table1).2.1 Read Alignment
and Quality Control
A single cell possesses only a very small amount of RNA, and the
sequencing reaction is limited by the amount of starting material.
Therefore, scRNA-seq experiments have several unique properties
including high technical noise [82], low coverage [116], and 3^0 bias
[117], requiring the use of methods different from bulk RNA-seqApplications of Single-Cell Sequencing for Multiomics 335