An effective and rapid method for RNA preparation from non-conventional yeast species

Abstract

The increased use of high-throughput RNA-based analysis has spurred the demand for rapid and simple preparation of high quality RNA. RNA preparation from non-conventional yeasts having diverse cell wall and morphological characteristics is often inefficient using current methods adapted for the model yeast, Saccharomyces cerevisiae. We report a simple RNA preparation method based on glass bead-mediated breakage in a formamide/EDTA solution. High quality RNA is generated within 15 min from various non-conventional yeasts species. The obtained RNA can be directly used for experimentation without further RNA purification and buffer exchange.

Full text

Contents lists available at ScienceDirect
Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio
An effective and rapid method for RNA preparation from non-conventional
yeast species
Dong Wook Lee
a
, Chang Pyo Hong
b
, Hyun Ah Kang
a,∗
a
Molecular Systems Biology Laboratory, Department of Life Science, Chung-Ang University, Seoul, 06974, South Korea
b
TheragenEtex Bio Institute, Suwon, 16229, South Korea
ARTICLE INFO
Keywords:
RNA extraction
Non-saccharomyces yeasts
Cell wall
Hyphae
ABSTRACT
The increased use of high-throughput RNA-based analysis has spurred the demand for rapid and simple pre-
paration of high quality RNA. RNA preparation from non-conventional yeasts having diverse cell wall and
morphological characteristics is often inefficient using current methods adapted for the model yeast,
Saccharomyces cerevisiae. We report a simple RNA preparation method based on glass bead-mediated breakage in
a formamide/EDTA solution. High quality RNA is generated within 15 min from various non-conventional yeasts
species. The obtained RNA can be directly used for experimentation without further RNA purification and buffer
exchange.
Advanced RNA-based analysis technologies, such as micro-array,
RNA sequencing, and quantitative real-time polymerase chain reaction
(qRT-PCR), require simple and rapid RNA preparation with minimal
exposure to unintended conditions to obtain sufficient quantities of
non-degraded, high quality RNA. Saccharomyces cerevisiae has been
used as a model yeast to study many characteristics of eukaryotes [1],
as an industrial microbe for many fermented beverages and bioethanol
production [2], and as the host strain for the production of useful
medicinal recombinant proteins and metabolites [3]. In recent decades,
non-conventional yeasts, or non-Saccharomyces yeasts species, have
drawn increased attention for diverse biotechnological applications [4].
Various non-conventional yeasts species are responsible for flavor de-
velopment in fermented foods like traditional alcoholic beverages [5],
and have unique characteristics, such as thermo/osmo-tolerance, that
are suitable for sustainable bioprocesses, such as simultaneous sac-
charification and fermentation for bioethanol production [6]. Several
non-conventional yeasts have been developed as a host system for
production of metabolites and recombinant proteins with industrial
potential [7]. Development of traditional molecular genetic techniques
and more advanced synthetic biology tools in non-conventional yeasts
species is expected to expand and diversify their impact on bio-
technology [8]. For a better understanding of the unique genomic and
functional characteristics of non-conventional yeasts species, whole-
genome sequencing and transcriptome analysis have become essential
to provide comprehensive information on physiological activities and
regulation of gene expression in metabolic pathways with potential
biotechnological importance. Comparative transcriptome profiling data
under diverse culture conditions will facilitate the identification of key
targets for metabolic engineering [9].
There are numerous genus and species of yeasts. They have very
different morphological characteristics that include capsule structure of
Cryptococcus neoformans [10], multi-polar hyphae of Saccharomycopsis
fibuligera [11],and pseudo-hyphae of Yarrowia lipolytica [12]. Isolation
of high quality total RNA from non-conventional yeasts with diverse
cell wall structure and morphological characteristics is often inefficient
using current methods and RNA extraction kits, which are adapted for
S. cerevisiae. Current methods to obtain RNA from S. cerevisiae include
acid hot-phenol extraction [13], a water bath method [14], the
RNAsnap™protocol [15], and one-step hot formamide extraction [16].
Although RNAs are efficiently obtained from S. cerevisiae using those
methods, the preparation of high quality RNA is often challenging in
other yeasts species that have mycelial formation with hyphae, thick
cell wall, and carbohydrate capsule structure. For total RNA extraction
from the multi-polar mycelial yeast S. fibuligera [11] and the en-
capsulated yeast C. neoformans [17], frozen yeast cells were ground in
liquid nitrogen using a mortar and a pestle before total RNA extraction.
However, grinding requires large amounts of cells and the RNA ob-
tained from the same quantity of yeast cells can vary markedly due to
the differences of physical force. We recently tried to prepare total RNA
samples from the non-conventional dimorphic pseudo-hyphal forming
yeast Hyphopichia burtonii [18] using several current RNA extraction
methods, including hot-acid phenol, grinding-combined with RNA
https://doi.org/10.1016/j.ab.2019.113408
Received 7 May 2019; Received in revised form 19 August 2019; Accepted 26 August 2019
∗
Corresponding author.
E-mail address: hyunkang@cau.ac.kr (H.A. Kang).
Analytical Biochemistry 586 (2019) 113408
Available online 27 August 2019
0003-2697/ © 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T

Fig. 1. Yeast RNA extraction method based on
FE/glass bead breakage. (A) Outline of the FE/
glass bead breakage method for small-scale RNA
extraction from yeast cells. (B) Gel analysis of RNA
samples in FE solution. H. burtonii and H. pseudo-
burtonii were cultivated in YPD with 1 M NaCl for
the indicated times (0–30 min) to impose an osmotic
stress. RNAs were extracted from the yeast cells
with total OD
600
2.0 and equal volumes of each
samples were separated by 1.2% agarose MOPS/
formaldehyde gel electrophoresis. (C) RNA sample
preparation from various non-conventional yeasts
species. Sf:S. fibuligera,Yl:Y. lipolytica,Wa:W.
anomalus,Ws:W. subpelliculosus,Op:O. para-
polymorpha, and Ca:C. albicans. Yeast cells were
cultivated in YPD and total RNAs were extracted
(total OD
600
2.0). (D) Optimization of glass bead
volume. Dh:D. hansenii,Cn:C. neoformans,Cn
Δcap59:C. neoformans acapsulular strain. Different
volumes of glass beads (0, 50, 100 μl) were added to
the yeast cells (OD
600
2.0) suspended in 100 μlFE
solution. RNA samples in FE solution were directly
mixed with 2 x RNA sample loading buffer (R1386-
1VL, Sigma-Aldrich) and SafePinky nucleotide
staining solution (S1001-025, GenDEPOT).
Fig. 2. Application of the RNA samples prepared by FE/glass bead breakage to downstream enzymatic reactions. Relative expression levels of H. burtonii and
H. pseudoburtonii genes (ENA5A and ERG11) upon 1 M NaCl treatment, as analyzed by qRT-PCR and RNA-Seq. The expression levels of genes by RNA-Seq were
quantified using TopHat and Cufflinks [24] with the value of fragments per kilobase of exon per million fragments mapped (FPKM), and differential expressions
between control (0 min) and other conditions (5, 15, or 30 min) were analyzed using Cuffdiffwith two replicates with the cutoffat p-value < 0.01. For qRT-PCR,
each sample was analyzed in duplicate and normalized by endogenous control genes HbATP4 and HpCYS3, respectively, which showed no apparent expression
changes at the transcript level in the presence of 1 M NaCl. The amplification efficiency of the primers of targets and control genes used for qRT-PCR was validated
very similar by analyzingㅿC
T
values with serially diluted cDNA.
D.W. Lee, et al. Analytical Biochemistry 586 (2019) 113408
2

isolation kit (Qiagen RNeasy Mini Kit, 74104), and one-step hot for-
mamide extraction. We failed to obtain high quality RNA (see Sup-
porting Fig. S1 A, B, and C in the supplementary material).
Here, we report a simple RNA extraction method based on glass
bead-mediated breakage in formamide/ethylenediaminetetraacetic
acid (EDTA) solution (FE) at room temperature (RT) (Fig. 1A). The
procedure is performed in microcentrifuge tubes and takes only 15 min.
1. Protocol
For RNA sample preparation, yeast cells are cultivated overnight in
2–3 ml of yeast extract peptone dextrose (YPD) medium (1% yeast ex-
tract, 2% peptone and 2% glucose) at 28 °C with shaking at 220 rpm.
The pre-cultured yeast cells are inoculated in fresh YPD medium at an
optical density at 600 nm (OD
600
) of 0.3–0.6, grown to an OD
600
of
approximately 1.0–2.0, which is nearly the early phase of exponential
growth.
Yeast cells (total 2 OD
600
unit) are transferred to 1.7 ml micro-
centrifuge tubes and harvested (13,000 rpm, 1 min, RT) using a tabletop
microcentrifuge. The supernatant is completely removed by pipetting
(an important step, since residual can affect the efficiency and quality
of the RNA extract) and the cells are washed by resuspension in 1 ml of
distilled water (DW) and centrifugation. The washed cells can be stored
at −80 °C after freezing with liquid nitrogen. When required, the cell
pellet is suspended in 100 μl of FE (98% formamide, 0.01 M EDTA),
prepared by mixing formamide (99.5%, F9037; Sigma-Aldrich) and
EDTA (0.5 M, pH 8.0, ML005-01; WelGENE). For optimal results, the FE
solution volume is increased in proportion to the volume of the cell
pellet (50 μl per 1.0 OD
600
of cell suspension). A 50 μl volume of RNase-
free glass beads (500 μm, GB05-RNA; Nextadvance) is added and the
cell suspension is ground by vortexing for three cycles of 30 s (total
90 s) at 6500 rpm at RT using a Pre-cellys 24 homogenizer (Bertin
Technologies). If a different homogenizer is used, such as an MT-360
TOMY (Seiko Corporation), the vortex time may need to be optimized
(see Supporting Fig. S1D,Fig. S1E, and Fig. S2). The optimal vortexing
time was defined as the shortest time generating the maximum quantity
of RNA with the 28S:18S ratio above 1.8, which was chosen as 90 s with
Pre-cellys 24 homogenizer in our study. The homogenized sample is
centrifuged at 13,000 rpm for 1 min at RT. The supernatant is trans-
ferred to a new 1.7 ml microcentrifuge tube. The extracted but un-
purified RNA in this FE solution can be directly assessed by RNA gel
electrophoresis in 10% 3-morpholinopropane-1-sulfonic acid (MOPS)/
0.75% formaldehyde running buffer. This RNA can also be directly used
for sequential DNase treatment or cDNA synthesis reaction. To avoid
interference of the enzyme reaction by formamide, it is critical to dilute
the RNA samples in FE solution to less than 5% of the total solution
[19]. If the RNA concentration in FE solution is higher than 100 ng/μl,
the RNA sample can be directly used for RNA-Seq analysis without any
further treatment. But, when the RNA concentration is not high enough,
ethanol precipitation is recommended to concentrate RNA. As an op-
tion, the RNA in FE solution can be diluted in diethyl pyrocarbonate
(DEPC)-treated water (C-9030; Bioneer). The sample can be stored at
−20 °C or −80 °C at this step.
The method yielded high quality RNA samples from the mycelial
yeast H. burtonii KJJ43 and H. pseudoburtonii KJS14 isolated from
Korean Nuruk fermentation starter (Supporting Table S1). The RNA
samples in FE solution were directly subjected to RNA formaldehyde-
agarose gel electrophoresis, which revealed non-degraded large rRNA
and low molecular weight tRNA and 5S rRNA (Fig. 1B). We further
applied this method to the preparation of RNA samples from various
non-conventional yeasts species, including S. fibuligera KJJ81, Y. lipo-
lytica PolF, Wickerhamomyces anomalus Y4, Wickerhamomyces sub-
pelliculosus SMY-04, Ogataea parapolymorpha DL1-L, Candida albicans
ATCC32354, Debaryomyces hansenii KD-2, Cryptococcus neoformans H99,
and C. neoformans acapsular mutant Δcap59 (Fig. 1C and D).In the case
of D. hansenii, which forms biofilms [20], and C. neoformans, which
possesses a thick extracellular capsule, we tested different volumes of
glass beads. The use of 50 μl was generally appropriate for the pre-
paration of RNAs from the most yeasts species tested in this study. An
increased amount of glass beads could produce more efficient cell lysis,
but could result in more degradation of RNA. Compared to the wild
type C. neoformans strain, RNA was extracted more efficiently from the
acapsular strain C. neoformans Δcap59, indicating that the capsule
structure might lessen yeast cell breakage by glass beads.
The quality of the RNA samples of H. burtonii and H. pseudoburtonii
cultivated with 1 M NaCl to impose osmotic stress was further assessed
by several quality control analyses after ethanol precipitation pur-
ification of the RNA samples in FE solution. The concentration and RNA
integrity number (RIN) of RNA samples prepared from small-scale yeast
cell suspensions (total OD
600
2.0) were sufficient for RNA-Seq analysis
(Supporting Tables S2A and B). In the case of H. burtonii, the final yields
of total RNA from the same total cell OD were variable among the
samples, due to the physiological change of this hyphae forming yeast
under salt stress conditions. Despite such differences in final yields
among the RNA samples from H. burtonii cells cultivated under different
osmotic conditions, such as cultivation in the presence of 1 M NaCl, 1 M
KCl, and 1 M sorbitol, the quality of RNA samples prepared by FE/glass
bead breakage method was mostly good enough for RNA-Seq analysis
with high ratio of 28S/18S and RIN values (Supporting Table S3). For
RNA-Seq analysis, the qualified RNA (1 μg) in each sample was sub-
jected to poly(A) mRNA enrichment by using magnetic beads with oligo
(dT) and then sheared into short fragments. Using reverse transcriptase
and random hexamer primers, the first strand cDNA of mRNA frag-
ments was synthesized, and the second strand cDNA was then synthe-
sized using DNA Polymerase I and RNase H. The synthesized cDNA was
subjected to end-repair and poly(A) tailing and connected with se-
quencing adapters using a TruSeq Stranded mRNA Sample Prep Kit
(Illumina). The proper cDNA fragments, purified by a BluePippin in-
strument (Sage Science), according to the manufacturer's instructions,
were selected for further PCR amplification. Subsequently, the libraries
were subjected to paired-end sequencing with a 100 bp read length
using an Illumina HiSeq 2500 platform. The feasibility of the RNA
samples of H. burtonii and H. pseudoburtonii for polymerase chain re-
action (PCR) analysis was assessed. The genes chosen for PCR were H.
burtonii ATP4 (HbATP4),encoding a mitochondrial ATP synthase sub-
unit, and H. pseudoburtonii CYS3 (HpCYS3),which codes for cy-
stathionine gamma-lyase in the cysteine assimilation pathway. These
genes were chosen because they were shown to be expressed con-
stitutively regardless of salt stress conditions in our RNA-Seq data (GEO
submission number GSE130141). The sequences of PCR primers are
presented in Supporting Table S4. When we conducted PCR with the
primer sets of HbATP4 and HpCYS3 using the RNA samples in FE so-
lution as templates, we did not observe any DNA bands amplified by
PCR before cDNA synthesis (Supporting Figs. S3A and B). The absence
of amplification of PCR products from the RNA samples without DNase
treatment indicated the lack of contamination by chromosomal DNA of
the RNA prepared by the FE/glass bead breakage method. For cDNA
synthesis, the RNA samples in FE solution were used directly without
buffer exchange and DNase treatment. After cDNA synthesis, we con-
firmed the amplification of HbATP4 and HpCYS3 DNA fragments by
PCR from the synthesized cDNAs (Supporting Figs. S3C and D).
To examine the consistency of expression patterns of H. burtonii and
H. pseudoburtonii genes between RNA-Seq data and other RNA analysis
data, we carried out qRT-PCR analysis with the gene specific primers
for ENA5A (encoding an ATPase sodium pump) and ERG11 (encoding
lanosterol 14-alpha-demethylase in the ergosterol biosynthesis
pathway) of H. burtonii and H. pseudoburtonii, using the synthesized
cDNAs as templates (Fig. 2).The activity of the sodium pump family is
important for osmotolerance [21], and the expression levels of genes
involved in the ergosterol pathway are decreased by osmotic stress in S.
cerevisiae [22]. Our RNA-Seq data revealed the increased expression of
ENA5A and the decreased expression of ERG11 in both H. burtonii and
D.W. Lee, et al. Analytical Biochemistry 586 (2019) 113408
3

H. pseudoburtonii, as expected. For cDNA synthesis, 1 μl of RNA in FE
buffer was diluted with 15 μl of DEPC-treated water, mixed with 4 μlof
SuperiorScript III Master Mix (Enzynomics) for cDNA synthesis reaction
at 42 °C for 45 min. The concentration of synthesized cDNA was ad-
justed to have Cq values within the range of 20–35 in qRT-PCR, which
was carried out with CFX96 Optical Module (Biorad) using TB Green
Premix Ex Taq (Takara). Gene expression values were calculated by the
2
−ΔΔCT
method [23] using HpCYS3 and HbATP4 as internal control,
respectively, for H. burtonii and H. pseudoburtonii genes in duplicated
experiments. The relative expression patterns of ENA5A and ERG11
detected by qRT-PCR were consistent with those based on RNA-Seq
data, although the fold change values were generally lower in the qRT-
PCR data compared to those in the RNA-Seq data (Fig. 2A and B), in-
dicating that this method ensures good quality of RNA samples, which
is important for experimentally reliable and repeatable results.
The RNA preparation method presented in this paper is a rapid
(within 15 min) and simple procedure that can be done at RT.
Compared to the widely used hot-acid phenol method and the recently
developed one-step hot formamide extraction method, this method does
not employ hot temperature and takes much less time to obtain high
quality RNA samples (Supporting Fig. S1F). With this method, a very
small amount of cell mass (OD
600
2.0) is sufficient to generate RNA
samples required for downstream applications, such as RNA-Seq and
qRT-PCR analysis. The method can be easily adapted for high-
throughput analysis of multiple RNA samples generated from diverse
yeasts species. With the increasing attention on non-conventional yeasts
species in both academic and industrial sectors, we anticipate that our
RNA isolation method will be useful for diverse RNA-based analyses for
a variety of yeasts species having diverse morphologies and cell wall
structures.
Acknowledgments
This work was supported by the National Research Foundation of
Korea, Grant No. NRF-2017M3C1B5019295 (STEAM Research Project)
and by the Korean Ministry of Agriculture, Food, and Rural Affairs,
Grant No. 918010042HD030 (Strategic Initiative for Microbiomes in
Agriculture and Food).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.ab.2019.113408.
References
[1] T. Bilinski, A. Bylak, R. Zadrag-Tecza, The budding yeast Saccharomyces cerevisiae as
a model organism: possible implications for gerontological studies, Biogerontology
18 (2017) 631–640.
[2] D.Q. Zheng, P.M. Wang, J. Chen, K. Zhang, T.Z. Liu, X.C. Wu, Y.D. Li, Y.H. Zhao,
Genome sequencing and genetic breeding of a bioethanol Saccharomyces cerevisiae
strain YJS329, BMC Genomics 13 (2012) 479–491.
[3] H. Kim, S.J. Yoo, H.A. Kang, Yeast synthetic biology for the production of re-
combinant therapeutic proteins, FEMS Yeast Res. 15 (2015) 1–16.
[4] R. Riley, S. Haridas, K.H. Wolfe, M.R. Lopes, C.T. Hittinger, et al., Comparative
genomics of biotechnologically important yeasts, Proc. Natl. Acad. Sci. U.S.A. 113
(2016) 9882–9887.
[5] A. Gamero, R. Quintilla, M. Groenewald, W. Alkema, T. Boekhout, L. Hazelwood,
High-throughput screening of a large collection of non-conventional yeasts reveals
their potential for aroma formation in food fermentation, Food Microbiol. 60 (2016)
147–159.
[6] S. Rebello, A. Abraham, A. Madhavan, R. Sindhu, P. Binod, K.B. Arun, E.M. Aneesh,
A. Pandey, Non-conventional yeast cell factories for sustainable bioprocesses, FEMS
(Fed. Eur. Microbiol. Soc.) Microbiol. Lett. 365 (2018) 1–10.
[7] V.T. Luu, H.Y. Moon, J.Y. Hwang, B.K. Kang, H.A. Kang, Development of re-
combinant Yarrowia lipolytica producing virus-like particles of a fish nervous ne-
crosis virus, J. Microbiol. 55 (2017) 655–664.
[8] J.M. Wanger, H.S. Alper, Synthetic biology and molecular genetics in non-con-
ventional yeasts: current tools and future advances, Fungal Genet. Biol. 89 (2016)
126–136.
[9] M. Bellasio, A. Peymann, M.G. Steiger, M. Valli, M. Sipiczki, M. Sauer, A.B. Graf,
H. Marx, D. Mattanovich, Complete genome sequence and transcriptome regulation
of the pentose utilizing yeast Sugiyamaella lignohabitans, FEMS Yeast Res. 16 (2016)
1–14.
[10] J.R. Perfect, Cryptococcus neoformans: a sugar-coated killer with designer genes,
FEMS Immunol. Med. Microbiol. 45 (2005) 395–404.
[11] J.H. Choo, C.P. Hong, J.Y. Lim, J.A. Seo, Y.S. Kim, D.W. Lee, S.G. Park, G.W. Lee,
E. Carroll, Y.W. Lee, H.A. Kang, Whole-genome de novo sequencing, combined with
RNA-Seq analysis, reveals unique genome and physiological features of the amy-
lolytic yeast Saccharomycopsis fibuligera and its interspecies hybrid, Biotechnol.
Biofuels 9 (2016) 246–268.
[12] J.C. Torres-Guzmán, A. Domínguez, HOY1, a homeo gene required for hyphal for-
mation in Yarrowia lipolytica, Molecular Cell Biology 17 (1997) 6283–6293.
[13] M.E. Schmitt, T.A. Brown, B.L. Trumpower, A rapid and simple method for pre-
paration of RNA from Saccharomyces cerevisiae, Nucleic Acids Res. 18 (1990)
3091–3092.
[14] J. Li, J. Liu, X. Wang, L. Zhao, Q. Chen, W. Zhao, A waterbath method for pre-
paration of RNA from Saccharomyces cerevisiae, Anal. Biochem. 384 (2009)
189–190.
[15] M.B. Stead, A. Agrawal, K.E. Bowden, R. Nasir, B.K. Mohanty, R.B. Meagher,
S.R. Kushner, RNAsnap: a rapid, quantitative and inexpensive, method for isolating
total RNA from bacteria, Nucleic Acids Res. 40 (2012) 1–9.
[16] D. Shedlovskiy, N. Shcherbik, D.G. Pestov, One-step hot formamide extraction of
RNA from Saccharomyces cerevisiae, RNA Biol. 12 (2016) 1722–1726.
[17] S.A. Cheon, E.J. Thak, Y.S. Bahn, H.A. Kang, A novel bZIP protein, Gsb1, is required
for oxidative stress response, mating, and virulence in the human pathogen
Cryptococcus neoformans, Sci. Rep. 7 (2017) 1–15.
[18] M. Groenewald, M.T. Smith, Re-examination of strains formerly assigned to
Hyphopichia burtonii, the phylogeny of the genus Hyphopichia, and the description of
Hyphopichia pseudoburtonii sp. nov, Int. J. Syst. Evol. Microbiol. 60 (2010)
2675–2680.
[19] P. Chomczynski, Solubilization in formamide protects RNA from degradation,
Nucleic Acids Res. 20 (1992) 3791–3792.
[20] F. Zhang, Y. Tang, Y. Ren, K. Yao, Q. He, Y. Wan, Y. Chi, Microbial composition of
spoiled industrial-scale Sichuan paocai and characteristics of the microorganisms
responsible for paocai spoilage, Int. J. Food Microbiol. 275 (2018) 32–38.
[21] M. Rep, M. Krantz, J.M. Thevelein, S. Hohmann, The transcriptional response of
Saccharomyces cerevisiae to osmotic shock. Hot1p and Msn2p/Msn4p are required
for the induction of subsets of high osmolarity glycerol pathway-dependent genes,
J. Biol. Chem. 275 (2000) 8290–8300.
[22] F.M. Montanes, A.P. Ahuir, M. Proft, Repression of ergosterol biosynthesis is es-
sential for stress resistance and is mediated by the Hog1 MAP kinase and the Mot3
and Rox1 transcription factors, Mol. Microbiol. 79 (2011) 1008–1023.
[23] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using
RealTime quantitative PCR and the 2
-ΔΔC
T
method, Methods 25 (2001) 402–408.
[24] C. Tranpnell, A. Roberts, L. Goff, G. Pertea, D. Kim, et al., Differential gene and
transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks,
Nat. Protoc. 7 (2012) 562–578.
D.W. Lee, et al. Analytical Biochemistry 586 (2019) 113408
4