Local Run Manager software. For other methods, please consult
the Illumina NextSeq 500 System Guide (Document
#15046563v06) and the NextSeq System Denature and Dilute
Libraries Guide (Document #15048776v09).
1.3 Computational
Component Overview
The complex nature of sequence analysis has led to the generation
of many tools and variations of these tools to address the evergrowing applications of sequencing technology. Thus, there are a
number of good choices for tools and pipelines for analysis. Here,
we will focus on the most common analyses used to process
M. tuberculosis RNAseq data with the ultimate goal of differential
gene expression analysis. We will also recommend the reader to use
the Galaxy platform (https://usegalaxy.org/) for analysis [7] (see
Note 1). In every case, the tool described can be used outside of
Galaxy, but the advantages of reproducibility, documentation, and
cloud-based computation have led us to recommend this platform
for most groups. For researchers with experience in bioinformatics,
it may be preferable to use each tool or some combination of tools
below without using the Galaxy platform. Assuming the use of
Galaxy’s main server for computation, there are still some files
and hardware that are required.
The workflow outlined below is accessible via a saved Galaxy
Workflow that is published and available at the link https://
usegalaxy.org/u/mark-jones/w/rnaseqbacteriamtb.
While the workflow can be used as is, some steps may need to
be performed individually or options changed for each tool to
ensure they are run properly.
The raw data generated from a sequencing run are stored in
files with the .FASTQ extension. These are text files that contain
the sequence of each read and the quality score associated with each
sequence. The workflow necessary to calculate gene expression
levels and determine differentially expressed genes from an RNAseq
dataset requires the steps summarized in Fig. 2 [8]. We have elected
to use Galaxy as a wrapper for these tools, but they may also be used
in their native form (typically R packages from Bioconductor [9]) if
that suits the user. First, the raw FASTQ data will need to undergo a
quality check, and sequenced reads will have base calls with low
confidence trimmed in silico (step 8).
Subsequently, reads are aligned to a reference genome of the
organism being profiled (step 9), which enables the assessment of
how many reads corresponding to each transcript in an organism’s
transcriptome were detected in each experiment (step 10). After
confirming the quality of the sequencing reads, the next step is to
align the reads to a reference genome or transcriptome that has
been previously generated. This step is accomplished using a
mapping tool that compares reads against a reference sequence.
Typically, the reference sequence will need to be transformed into
a genome index, which is a data structure that is more amenable to
486
Shuyi Ma et al.
the Illumina NextSeq 500 System Guide (Document
#15046563v06) and the NextSeq System Denature and Dilute
Libraries Guide (Document #15048776v09).
1.3 Computational
Component Overview
The complex nature of sequence analysis has led to the generation
of many tools and variations of these tools to address the evergrowing applications of sequencing technology. Thus, there are a
number of good choices for tools and pipelines for analysis. Here,
we will focus on the most common analyses used to process
M. tuberculosis RNAseq data with the ultimate goal of differential
gene expression analysis. We will also recommend the reader to use
the Galaxy platform (https://usegalaxy.org/) for analysis [7] (see
Note 1). In every case, the tool described can be used outside of
Galaxy, but the advantages of reproducibility, documentation, and
cloud-based computation have led us to recommend this platform
for most groups. For researchers with experience in bioinformatics,
it may be preferable to use each tool or some combination of tools
below without using the Galaxy platform. Assuming the use of
Galaxy’s main server for computation, there are still some files
and hardware that are required.
The workflow outlined below is accessible via a saved Galaxy
Workflow that is published and available at the link https://
usegalaxy.org/u/mark-jones/w/rnaseqbacteriamtb.
While the workflow can be used as is, some steps may need to
be performed individually or options changed for each tool to
ensure they are run properly.
The raw data generated from a sequencing run are stored in
files with the .FASTQ extension. These are text files that contain
the sequence of each read and the quality score associated with each
sequence. The workflow necessary to calculate gene expression
levels and determine differentially expressed genes from an RNAseq
dataset requires the steps summarized in Fig. 2 [8]. We have elected
to use Galaxy as a wrapper for these tools, but they may also be used
in their native form (typically R packages from Bioconductor [9]) if
that suits the user. First, the raw FASTQ data will need to undergo a
quality check, and sequenced reads will have base calls with low
confidence trimmed in silico (step 8).
Subsequently, reads are aligned to a reference genome of the
organism being profiled (step 9), which enables the assessment of
how many reads corresponding to each transcript in an organism’s
transcriptome were detected in each experiment (step 10). After
confirming the quality of the sequencing reads, the next step is to
align the reads to a reference genome or transcriptome that has
been previously generated. This step is accomplished using a
mapping tool that compares reads against a reference sequence.
Typically, the reference sequence will need to be transformed into
a genome index, which is a data structure that is more amenable to
486
Shuyi Ma et al.