Biology Reference
In-Depth Information
up to 210 kb upstream of the TSS annotated in XenBase or Ensembl data-
bases. A limiting factor of conventional RNA-Seq is related to the length of
the sequence reads (
50 bp), that can be difficult to map when they overlap
over two exons or when exons are small. Because of this, conventional
RNA-Seq does not allow precise definition of exon boundaries and is
not well suited for the definition of genes structure. This technical limitation
is overcome with longer reads of 75-100 bp, and pair-end sequencing,
which allows a better resolution when defining exons boundaries
(
Fig. 10.3
B) (see below). This technology has been successfully used to re-
fine current gene models in
X. tropicalis
(
Akkers et al., 2009
).
Several recent protocols allow longer sequence reads (
100 bp) and
pair-end sequencing, in which the two ends of each cDNA fragment are
sequenced. Although the analysis is more complex, these protocols are well
suited for
de novo
transcript reconstruction, exon definition, detection, and
quantitative measure of alternative splicing. Two strategies can be used to
process the sequenced reads (
Garber et al., 2011
): The reads can be assem-
bled into transcripts, including splicing variants, which can be mapped to a
reference genome, thus providing exon annotation. Alternatively, the reads
can be mapped directly to the reference genome, without
de novo
assembly,
and alternative splicing can be inferred from reads split in two and mapping
over the ends of two distinct exons (also known as “spliced reads”). These
methods provide a single base pair resolution of exon boundaries.
The 5
0
end of transcripts are often missed and are poorly annotated. To
this regard, the RNA-PET technology holds many promises. Originally de-
veloped to detect fusion transcripts in cancer cells, the RNA-PET specifi-
cally captures the 5
0
and the 3
0
ends of transcripts (
Ruan &Ruan, 2011
). The
principle of this method is depicted in
Fig. 10.4
A. Full-length cDNA is syn-
thesized from mRNA transcripts with CAP-Trapper and GsuI-dT(16) pu-
rification. A linker containing the binding site of the Mme1 (or
Eco
P15I)
endonuclease is then ligated to the full-length cDNA, allowing it to be cir-
cularized. The circular DNA molecule is then restricted with
Mme
1, which
cuts 27 bp downstream of its binding site, thus resulting in 27 bp tags of each
end of the original transcript, spaced by the linker. Sequencing adaptors are
then ligated for high-throughput sequencing. The resulting paired-end tags
(PETs) are then mapped to the genome, thus providing experimental evi-
dence for the 5
0
and 3
0
ends of transcripts, including potential alternative TSS
and transcription termination sites (TTSs), but provides no information rel-
ative to the their internal structure (
Fig. 10.4
B). Thus, RNA-PET provides
complementary information about the boundaries of the transcriptional
