Page 111 - Read Online

Page 245 Ratnapriya. J Transl Genet Genom 2022;6:240-256 https://dx.doi.org/10.20517/jtgg.2021.54

METHODS OF GENE EXPRESSION MEASUREMENT
Advent in next-generation sequencing has led to major advances in methods for gene expression profiling.
In the last few years, RNA-sequencing (RNA-seq) has become the most common method of transcriptome
profiling because of precise, quantitative measurement. Broadly, RNA-seq experiments can be divided into
two categories:

1. Bulk RNA-seq: This method measures the gene expression from heterogeneous tissues for detection of a
wide variety of RNA species, including mRNA, non-coding RNA, transcript isoforms, and circulating and
[74]
small regulatory RNA . It is relatively fast and inexpensive, and in the past decade, it has been widely used
to investigate multiple aspects of biology and biomedical research including connecting GWAS findings to
[75]
function . Specifically, it has greatly facilitated in characterizing the molecular function of the human
genome and the impact of genetic variation on gene expression levels . Additionally, profiling of
[51]
transcriptomes in large cohorts of cases and controls also presents an opportunity for identifying genes and
pathways affected by the disease, enabling mechanistic insights into disease. However, as tissue is a mix of
various cell-types of varying proportions, the bulk RNA-seq averages the expression across all cells in a
sample, which can significantly confound the analysis of differential gene expression and eQTL analyses.

2. Single-cell RNA sequencing (scRNA-seq): This method allows the transcriptome profiling at the single-
cell level, which can address many technical and biological limitations associated with bulk RNA-seq
discussed above. This has great potential for providing insights into the understanding of biological
diversity and rare cell-types that could not be resolved using bulk RNA-seq. scRNA-seq has been extensively
applied for cell-types identification, classification and lineage tracing . Cell-type-specific transcriptome,
[76]
combined with variant information also offers an opportunity to implicate cell-type specificity of traits and
diseases. In addition, this approach increases the power to detect such associations as expression profile is
less heterogeneous for single-cell than bulk tissues. However, scRNA-seq resolution is often noisy, sparse,
and high-dimensional, creating challenges for computational analysis. Additionally, it is still relatively
expensive to apply to a large cohort for gaining insights into the cell types affected by the disease.

LEVERAGING TRANSCRIPTOMICS FOR IDENTIFICATION OF CAUSAL VARIANT AND
TARGET GENE
Linking the AMD-GWAS variants to their target genes is an important step in understanding the role of the
non-coding genome in disease. This can be done in two ways: firstly, by statistical fine-mapping through
conditional association analysis that can help identify the credible variants that are most likely to be causal,
and variants can be linked to causal genes through functional analysis . However, identifying the causal
[77]
SNP is complicated because of LD. Alternatively, GWAS loci can be analyzed for eQTLs [78,79] mapping where
genotypes are correlated with the gene expression. They are divided into two classes: cis-eQTLs and trans-
eQTLs [Figure 2]. eQTLs that affect the expression of transcripts from nearby genes (generally within 1 MB
of transcription start site) are referred to as cis-eQTLs. These variants often regulate gene expression by
influencing the transcription factor binding sites that can impact multiple levels of gene expression and
chromatin organization such as histone modification and DNA methylation. cis-eQTLs are abundant in all
species including humans. Across all tissues, 94.7% of all protein-coding and 67.3% of all long-non coding
RNA (lincRNA) genes have an eQTL , whereas, in the retina, 81% of known protein-coding and 13% of
[51]
non-coding genes are under genetic regulation . Trans-eQTLs affect the expression of distant genes either
[80]
on the same chromosome or elsewhere in the genome. Not much is known regarding the effects of trans-
eQTLs in humans as they have a small effect size and a large number of samples need to be analyzed to
[81]
discover trans effects. They are also far more variable across species and tissue-specific than cis-eQTLs .