1. RNA-Seq and Transcriptome Analysis
Jessica R. Holmes
Research & Instructional Specialist in Life Sciences
High Performance Biological Computing (HPCBio)
Roy J. Carver Biotechnology Center

2. General Outline Getting the RNA-Seq data: from RNA -> Sequence data
Experimental and practical considerations
Commonly encountered file formats
Transcriptomic analysis methods and tools
Transcriptome Assembly
Differential Gene expression

3. Transcriptome Sequencing (aka RNA-Seq)
Why sequence RNA?
Differential Gene Expression
Quantitative evaluation and comparison of transcript levels, usually between different groups
Vast majority of RNA-Seq is for DGE
Transcriptome Assembly
Build new or improved profile of transcribed regions (“gene models”) of the genome
Can then be used for DGE
Metatranscriptomics
Transcriptome analysis of a community of different species (e.g., gut bacteria, hot springs, soil)
Gain insights on the functioning and activity rather than just who is present

4. Types of RNA Ribosomal (rRNA) Messenger (mRNA ) Micro (miRNA)
Responsible for protein synthesis
60% of a ribosome; up to 95% of total RNA in a cell
Messenger (mRNA )
Translated into protein in ribosome
3-4% of total RNA in a cell
have poly-A tails in eukaryotes
Micro (miRNA)
short (22 bp) non-coding RNA involved in expression regulation
Transfer (tRNA)
Bring specific amino acids for protein synthesis
Others (lncRNA, shRNA, siRNA, snoRNA, etc.)

5. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Usually don’t want to sequence rRNA, so need to select for other RNA types by:
poly-A selection (eukaryotes only)
ribosomal depletion
Size selection

6. Illumina Sequencing Workflow
Fragment DNA
Repair ends
Add A overhang
Ligate adapters
Purify
Library Preparation 1 6

7. From RNA -> sequence data
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

8. From RNA -> sequence data
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

9. From RNA -> sequence data
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

10. How do we sequence DNA? 1st generation: Sanger method (1987)
2nd generation (“next generation”; 2005):
454 - pyrosequencing
SOLiD – sequencing by ligation
Illumina – sequencing by synthesis
Ion Torrent – ion semiconductor
Pac Bio – Single Molecule Real-Time sequencing, 1000 bp
3rd generation (2015)
Pac Bio – SMRT, Sequel system, 20,000 bp
Nanopore – ion current detection
10X Genomics – novel library prep for Illumina

11. Illumina – “short read” sequencing
Rapid improvements over the years from 36 bp to 300 bp; highest throughput at 100 bp; many different types of sequencers for various applications.
Can also “flip” a longer DNA strand and sequence from the other end to get paired-end reads
Accuracy: 99.99% Biases: yes
Most common platform for transcriptome sequencing
……………..
Single-read
100nt
Paired-end

12. Illumina Sequencing Workflow
Fragment DNA
Repair ends
Add A overhang
Ligate adapters
Purify
Library Preparation 1
Cluster Generation
Hybridize to flow cell
Extend hybridized template
Perform bridge amplification
Prepare flow cell for sequencing 2
Sequencing
Perform sequencing Generate base calls 3
Data Analysis
Images
Intensities
Reads
Alignments 4 12

13. Illumina Sequencing Technology Workflow
Single molecule array
Cluster Growth
Sequencing 5’ 3’
DNA ( μg) C A T G T G T A C
Library Preparation
Image Acquisition 1 2 3 7 8 9 4 5 6
Base Calling
T G T A C G A T … 13

14. Illumina Sequencing Video

15. General Outline Getting the RNA-Seq data: from RNA -> Sequence data
Experimental and Practical considerations
Commonly encountered file formats
Transcriptomic analysis methods and tools
Transcriptome Assembly
Differential Gene expression

16. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Experimental design
Technical replicates
Illumina has low technical variation unlike microarrays
Technical replicates are unnecessary
Batch effects
Best to sequence everything for an experiment at the same time
If you are preparing the libraries, be consistent & make them simultaneously
Biological replicates
This is essential for your experiment to have any statistical power
At least 3, but the more the better

17. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Experimental design
For transcriptome assembly
RNA can be pooled from various sources to ensure the most robust transcriptome
Pooling can also be done after sequencing, but before assembly
For differential gene expression
Pooling RNA from multiple biological replicates is usually not advisable

18. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Poly(A) enrichment or ribosomal RNA depletion?
Unless interested in rRNA, it’s always best to remove
Remember it makes up ~ 95% of the transcriptome
Use Poly(A) enrichment if you are only interested in coding RNA
This is very common, but do not use if interested in non- coding RNA as well
For metatranscriptomics, you should also remove host DNA, but this can be done with bioinformatics tools downstream

19. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Single-end (SE) or Paired end (PE)?
SE is most common for DGE analysis
PE is best for assemblies, for isoform differentiation, and for paralogous & orthologous gene differentiation (i.e. high-ploidy genomes & metatransciptomes)
Single-end read
Read1
ATGTTCCATAAGC…
Paired-end reads
Read1
ATGTTCCATAAGC…
Read2
CCGTAATGGCATG…

20. Experimental and Practical considerations
RNA-Seq
Experimental and Practical considerations
Stranded?
Most RNA-Seq library preparation kits produce stranded libraries
Exception: single cell prep kits may be unstranded
Can identify which strand of DNA the RNA was transcribed from
Strandedness is advisable for all applications
Three types of libraries
Unstranded – The strand of DNA used to transcribe the reads is unknown
Reverse – Reads were transcribed from the strand with complementary sequence to the reads
Forward – Reads were transcribed from the strand that has a sequence identical to the reads

21. From RNA -> sequence data
Uracil DNA Glycosylase
Uracil DNA Glycosylase
Borodina T., Methods in Enzymology (2011) 500:79–98

22. General Outline Getting the RNA-Seq data: from RNA -> Sequence data
Experimental and Practical considerations
Commonly encountered file formats
Transcriptomic analysis methods and tools
Transcriptome Assembly
Differential Gene expression

23. File formats Sequence formats FASTA FASTQ Feature formats GFF GTF
A brief note
Sequence formats
FASTA
FASTQ
Feature formats
GFF
GTF
Alignment formats
SAM
BAM

24. Formats: FASTA Deceptively simple format (e.g. there is no standard)
>unique_sequence_ID My sequence is pretty cool
ATTCATTAAAGCAGTTTATTGGCTTAATGTACATCAGTGAAATCATAAATGCTAAAAA
Deceptively simple format (e.g. there is no standard)
However in general:
Header line, starts with ‘>’
followed directly by an ID
… and an optional description (separated by a space)
Files can be fairly large (whole genomes)
Any residue type (DNA, RNA, protein), but simple alphabet

25. Formats: FASTA E.g. a read E.g. a chromosome >unique_sequence_ID
ATTCATTAAAGCAGTTTATTGGCTTAATGTACATCAGTGAAATCATAAATGCTAAAAATTTATGATAAAA
>Group10 gi| |ref|NC_ | Amel_4.5, whole genome shotgun sequence
TAATTTATATATCTATTTTTTTTATTAAAAAATTTATATTTTTGTTAAAATTTTATTTGATTAGAAATAT
TTTTACTATTGTTCATTAATCGTTAATTAAAGATAGCACAGCACATGTAAGAATTCTAGGTCATGCGAAA
TTAAAAATTAAAAATATTCATATTTCTATAATAATTAAATTATTGTTTTAATTTAAGTAAAAAAATTTCT
AAGAAATCAAAAATTTGTTGTAATATTGAAACAAAATTTTGTTGTCTGCTTTTTATAGTAACTAATAAAT
ATTTAATAAAAAATTACTTTATTTAATATTTTATAATAAATCAAATTGTCCAATTTGAAATTTATTTTAT
CACTAAAAATATCTTTATTATAGTCAATATTTTTTGTTAGGTTTAAATAATTGTTAAAATTAGAAAATGA
TCGATATTTTCAAATAGTACGTTTAACTAATACTTAAGTGAAAGGTAAAGCGGTTATTTAAAATATTGAT
TTATAATATTCGTGACATAATATATTTATAAATAGATTATATATATATATATACATCAAAATATTATACG
AGAACTAGAAAATATTACAGATGCAAAATAAATTAAATTTTGTAAATGTTACAGAATTAAAAATCGAAGT

26. Formats: FASTQ FASTQ – FASTA with quality
@unique_sequence_ID
ATTCATTAAAGCAGTTTATTGGCTTAATGTACATCAGTGAAATCATAAATGCTAAAAATTTATGATAAAA +
DNA sequence with quality metadata
The header line, starts with directly by an ID and an optional description (separated by a space)
May be ‘raw’ data (straight from sequencing) or processed (trimmed)
Variations: Sanger, Illumina, Solexa (Sanger is most common)
Can hold 100’s of millions of records
Files can be very large - 100’s of GB apiece

27. “Phred” quality (Q) scores
Whole-genome sequencing and comprehensive variant analysis of a Japanese individual using massively parallel sequencing
Akihiro Fujimoto, Hidewaki Nakagawa, Naoya Hosono, Kaoru Nakano, Tetsuo Abe, Keith A Boroevich, Masao Nagasaki, Rui Yamaguchi, Tetsuo Shibuya, Michiaki Kubo, Satoru Miyano, Yusuke Nakamura & Tatsuhiko Tsunoda
Nature Genetics, 2010
Q score
Prob. of wrong call
Accuracy 10
1 in 10 (0.1)
90% 20
1 in 100 (0.01)
99% 30
1 in 1000 (0.001)
99.9% 40
1 in (0.0001)
99.99%

28. Formats: FASTQ Sanger = Illumina 1.8+ FASTQ – FASTA with quality
@unique_sequence_ID
ATTCATTAAAGCAGTTTATTGGCTTAATGTACATCAGTGAAATCATAAATGCTAAAAATTTATGATAAAA+unique_sequence_ID
Sanger = Illumina 1.8+

29. Feature formats
GTF/GFF3
SAM/BAM
UCSC formats (BED, WIG, etc.)

30. Feature formats
Used for mapping features against a particular sequence or genome assembly
May or may not include sequence data
The reference sequence must match the names from a related file (possibly FASTA)
These are version (assembly)-dependent - they are tied to a specific version (assembly/release) of a reference genome
Not all reference genomes are the represented the same! E.g. human chromosome 1
UCSC – ‘chr1’
Ensembl/NCBI – ‘1’
Best practice: get these from the same source as the reference

31. Feature formats : GTF Gene transfer format
Apr-18
Feature formats : GTF Gene transfer format
Differences in representation of information make it distinct from GFF
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan start_codon gene_id "001"; transcript_id "001.1";
AB Twinscan stop_codon gene_id "001"; transcript_id "001.1";
Source
End location
Strand
Attributes (hierarchy)
Chromosome ID
Start location
Reading frame
Gene feature
Score (user defined)

32. Feature formats : GTF Gene transfer format
Apr-18
Feature formats : GTF Gene transfer format
Differences in representation of information make it distinct from GFF
Source of GTF is important – Ensembl GTF is not quite the same as UCSC GTF
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan start_codon gene_id "001"; transcript_id "001.1";
AB Twinscan stop_codon gene_id "001"; transcript_id "001.1";
*********Could add this last point to previous slide
Source
End location
Strand
Attributes (hierarchy)
Chromosome ID
Start location
Reading frame
Gene feature
Score (user defined)

33. Feature formats : GFF3 Gene feature format (v3)
Tab-delimited file to store genomic features, e.g. genomic intervals of genes and gene structure
Meant to be unified replacement for GFF/GTF (includes specification)
All but UCSC have started using this (UCSC prefers their own internal formats)
Chr1 amel_OGSv gene ID=GB42165
Chr1 amel_OGSv mRNA ID=GB42165-RA;Parent=GB42165
Chr1 amel_OGSv ’UTR Parent=GB42165-RA
Chr1 amel_OGSv exon Parent=GB42165-RA
Chr1 amel_OGSv exon Parent=GB42165-RA
Source
End location
Strand
Attributes (hierarchy)
Chromosome ID
Start location
Gene feature
Phase
Score (user defined)

34. Feature formats: GFF3 vs. GTF
Apr-18
Feature formats: GFF3 vs. GTF
GFF3 – Gene feature format
GTF – Gene transfer format
Always check which of the two formats is accepted by your application of choice, sometimes they cannot be swapped
Chr1 amel_OGSv gene ID=GB42165
Chr1 amel_OGSv mRNA ID=GB42165-RA;Parent=GB42165
Chr1 amel_OGSv ’UTR Parent=GB42165-RA
Chr1 amel_OGSv exon Parent=GB42165-RA
Chr1 amel_OGSv exon Parent=GB42165-RA
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan CDS gene_id "001"; transcript_id "001.1";
AB Twinscan start_codon gene_id "001"; transcript_id "001.1";
AB Twinscan stop_codon gene_id "001"; transcript_id "001.1";

35. Apr-18
What is an alignment?
Wikipedia - “a way of arranging the sequences of DNA, RNA, or protein to identify regions of similarity that may be a consequence of functional, structural, or evolutionary relationships between the sequences”
How can we store this information about millions of reads that align to our reference genome?

36. Formats : SAM SAM – Sequence Alignment/Map format
SAM file format stores alignment information
Plain text
Specification:
Contains quality information, meta data, alignment information, sequence etc.
Files can be very large: Many 100’s of GB or more
Normally converted into BAM to save space (and text format is mostly useless for downstream analyses)
@HD [format version]
@SQ SN:chr_1 LN:
@PG [information about program that made this]
HWI-D00758:59:C7U2JANXX:1:1101:1398: chr_ S9M * NAGCTCTTTA #/<<BFBBFF NH:i:1 HI:i:1 AS:i:93 nM:i:2

37. Formats : BAM BAM – BGZF compressed SAM format
Compressed/binary version of SAM and is not human readable. Uses a specialized compression algorithm optimized for indexing and record retrieval (bgzip)
Makes the alignment information easily accessible to downstream applications (large genome file not necessary)
Unsorted, sorted by sequence name, sorted by genome coordinates
May be accompanied by an index file (.bai) (only if coordinate sorted)
Files are typically very large: ~ 1/5 of SAM, but still very large

38. General Outline 4. Transcriptomic analysis methods and tools
Transcriptome Analysis; aspects common to both assembly and differential gene expression
Download data
Quality check
Data alignment
Assembly
Differential Gene Expression
Choosing a method, the considerations…
Final thoughts and observations

39. Obtain sequence data
If you are using the R.J.C. Biotechnology Center and the Biocluster
Globus is most direct route
CNRG instructions
Download data to a computer and upload to Biocluster using an SFTP client
Cyberduck, WinSCP…
Can also use linux commands such as:
scp, rsync, wget, …

40. Globus

41. Quality Checks How do my newly obtained data look?
Check for overall data quality. FastQC is a great tool that enables the quality assessment.
Good quality!
Poor quality!

42. Quality Checks How do my newly obtained data look?
Check for overall data quality. FastQC is a great tool that enables the quality assessment.
In addition to the quality of each sequenced base, it will give you an idea of
Presence of, and abundance of contaminating sequences.
Average read length
GC content
NOTE – FastQC is good, but it is very strict and will not hesitate to call your dataset bad on one of the many metrics it tests the raw data for. Use logic and read the explanation for why and if it is acceptable.

43. Quality Checks What do I do when FastQC calls my data poor?
Poor quality at the ends can be remedied
“quality trimmers” like trimmomatic, fastx-toolkit, etc.
Left-over adapter sequences in the reads can be removed
“adapter trimmers” like trimmomatic.
Always trim adapters as a matter of routine
We need to amend these issues so we get the best possible alignment
After trimming, it is best to rerun the data through FastQC to check the resulting data

44. Transcriptome Analysis
Quality Checks
Before quality trimming
After quality trimming

45. Transcriptome Analysis
Data Alignment
We need to align the sequence data to our genome of interest
If aligning RNASeq data to the genome, almost always pick a splice-aware aligner
Genome
Gene
Reads
Versus
Alignment
Splice-Aware

46. Transcriptome Analysis
Data Alignment
We need to align the sequence data to our genome of interest
If aligning RNA-Seq data to the genome, always pick a splice-aware aligner (unless it’s a bacterial genome!)
HiSat2, TopHat2, STAR, MapSplice, SOAPSplice, Passion, SpliceMap, RUM, ABMapper, CRAC, GSNAP, HMMSplicer …
There are excellent aligners available that are not splice-aware. This is ideal for bacterial genomes.
BWA, Novoalign (not free), Bowtie2, SOAPaligner, HiSat2

47. Transcriptome Analysis
Data Alignment
Other considerations when choosing an aligner:
How does it deal with reads that map to multiple locations?
How does it deal with paired-end versus single-end data?
How many mismatches will it allow between the genome and the reads?
What assumptions does it make about my genome, and can I change these assumptions?

48. Transcriptome Analysis
Data Alignment
Differential gene expression
IGV is the visualization tool used for this snapshot

49. General Outline 4. Transcriptomic analysis methods and tools
Transcriptome Analysis; aspects common to both assembly and differential gene expression
Download data
Quality check
Data alignment
Assembly
Differential Gene Expression
Choosing a method, the considerations…
Final thoughts and observations

50. Transcriptome Assembly Overview
Two main types of assembly
Reference-based assembly
A de novo assembly
Differential gene expression

51. Transcriptome Assembly
Reference-based assembly
Used when the genome reference sequence is known, and:
Transcriptome data is not available
Transcriptome data is available but not good enough,
i.e. missing isoforms of genes, or unknown non-coding regions
The existing transcriptome information is for a different tissue type
Stringtie, Cufflinks and Scripture are two reference-based transcriptome assemblers

52. Transcriptome Assembly
Reference-based assembly
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

53. Transcriptome Assembly
Reference-based assembly
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

54. Transcriptome Assembly
Reference-based assembly
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

55. Transcriptome Assembly
Reference-based assembly
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

56. Transcriptome Assembly
De novo assembly
Used when very little information is available for the genome
Often the first step in putting together information about an unknown genome
Amount of data needed for a good de novo assembly is higher than what is needed for a reference-based assembly
Can be used for genome annotation, once the genome is assembled
Trinity, and TransABySS, are examples of well-regarded transcriptome assemblers
Transcriptome assembly, de novo, reference-based

57. Transcriptome Assembly
De novo assembly (De Bruijn graph construction)
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

58. Transcriptome Assembly
De novo assembly (De Bruijn graph construction)
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

59. Transcriptome Assembly
De novo assembly (De Bruijn graph construction)
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

60. Combined Transcriptome Assembly
Align-then-assemble preferred for a trusted reference genome & for filtering out pathogen RNA
Assemble-then-align preferred when reference genome is not trusted or is a closely related species to the reference
Martin J.A. and Wang Z., Nat. Rev. Genet. (2011) 12:671–682

61. How good is my assembly? Are all the genes I expected in the assembly?
Do I have complete genes?
Are the contigs assembled correctly?
How does it look compared to a close reference?

62. Tools for Evaluating Assembly: using the information you have
TransRate – evaluates assembly using reads, paired end information, reference genome, protein data, etc.
Can generate a ‘cleaned-up’ or optimized assembly based on metrics
DETONATE (from the RSEM developers) – evaluates assembly based on read mapping and/or reference information

63. Tools for Evaluating Assembly: conserved gene sets
CEGMA : From Ian Korf’s group, UC Davis Mapping Core Eukaryotic Genes
- Discontinued!
BUSCO: From Evgeny Zdobnov’s group, University of Geneva Coverage is indicative of quality and completeness of assembly

64. Outline Transcriptomic analysis methods and tools
Transcriptome Analysis; aspects common to both assembly and differential gene expression
Quality check
Data alignment
Assembly
Differential Gene Expression
Choosing a method, the considerations…
Final thoughts and observations

65. Differential Gene Expression Overview
Obtain/download sequence data
Check quality of data and
Trim low quality bases, and remove adapter sequence
Align trimmed reads to genome of interest
Pick alignment tool
Index genome file
Run alignment after choosing the relevant parameters
If you are not studying human or mouse, check every parameter and confirm that they make the correct assumptions for your genome! Otherwise, change them
Differential gene expression

66. Differential Gene Expression overview
Set up to do differential gene expression (DGE)
Identify read counts associated with genes
Do you want to obtain raw read counts or normalized read counts? This will depend on the statistical analysis you wish to perform downstream
htseq & feature-counts return raw read counts
Required for R programs like DESeq & EdgeR
Ballgown & Cufflinks return FPKM normalized counts for each gene
Differential gene expression

67. Differential Gene Expression
Options for DGE analysis

68. Differential Gene Expression
Options for DGE analysis

69. Differential Gene Expression
Options for DGE analysis

70. DGE Statistical Analyses
The first step is proper normalization of the data
Often the statistical package you use will have a normalization method that it prefers and uses exclusively (e.g. Voom, FPKM, TMM (used by EdgeR))
Is your experiment a pairwise comparison?
Ballgown/Cufflinks, EdgeR, DESeq
Is it a more complex design?
EdgeR, DESeq, other R/Bioconductor packages

71. Outline Transcriptomic analysis methods and tools
Transcriptome Analysis; aspects common to both assembly and differential gene expression
Download data
Quality check
Data alignment
Assembly
Differential Gene Expression
Choosing a method, the considerations…
Final thoughts and observations

72. Transcriptome Analysis
How does one pick the right tools?

73. What does HPCBio use? Quality Check - FASTQC Trimming - Trimmomatic
Splice-aware alignment - STAR
Bacterial alignment - BWA or Novoalign
Counting reads per gene - featureCounts
DGE Analysis - EdgeR or DESeq
De novo transcriptome assembly - Trinity & RSEM

74. You won’t be using these tools today 
Takes a lot of time to learn
We do offer a long workshop on these methods during Spring break every year
Check at the beginning of the year for updates

75. For today’s lab, you will use Tuxedo Suite
Easier and faster to learn
Requires only a basic background in Linux (maybe R)
Available on Galaxy, an online set of tools that’s ideal for small projects and teaching!
Great option for simple statistical models
Can perform all steps for DGE analysis very quickly
Can do splice-aware & regular alignments (no need to switch software for bacteria)
Can perform reference-based transcriptome assembly
Does DGE analysis and creates statistical graphs

76. Why doesn’t HPCBio use Tuxedo suite?
Very new. Has only been tested/optimized with human data
We work on a wide variety of organisms, and STAR works very well for us. No need to change yet…
Can’t handle complex statistical models
No more complicated than a pairwise-comparison
We need more flexibility
Doesn’t output raw read counts easily
This is needed for use of R packages like DESeq & EdgeR, which can handle complex models

77. Differential Gene Expression
Bowtie/Bowtie2 use Burrows-Wheeler indexing for aligning reads (not splice-aware)
Tophat/Tophat2 utilizes either Bowtie or Bowtie2 to align reads in a splice-aware manner and aids the discovery of new splice junctions
The Cufflinks package has 4 components, the 2 major ones are listed below:
Cufflinks does reference-based transcriptome assembly
Cuffdiff does DGE analysis in a simple pairwise comparison, and a series of pairwise comparisons in a time-course experiment
Options for DGE analysis (tuxedo suite)
Trapnell et al., Nature Protocols, March 2012

78. Old Tuxedo suite NEW Tuxedo suite
Bowtie/Bowtie2 - general aligner
Tophat/Tophat2 - splice-aware aligner
Cufflinks suite - Transcriptome assembly, gene counting, and DGE calculation for simple models
This suite is well-known, but there are a few issues with methods used
No longer being supported!
HISAT2 - DNA & RNA aligner in one
StringTie - Transcriptome assembly, & gene counting
Ballgown - Improved DGE calculation for simple models, and visualization
This suite is very new and attempts to improve on the old suite’s issues
However, not well-tested on anything but human at the moment

79. Old Tuxedo Suite Protocol
.fastq
.bam
.gtf or .gff3
Text
A single merged gtf
Trimmed sequence FASTQ files
Alignment file
Gene annotation file
.gtf or .gff3
Alignment file
Trapnell et al., Nature Protocols, March 2012

80. New Tuxedo Suite Protocol
Trimmed sequence FASTQ files
Alignment file
Gene annotation file
Pertea, et al., Nature Protocols, August 2016

81. Note about today’s lab
Unfortunately, the new Tuxedo suite isn’t fully available on Galaxy
Hopefully this will change soon
We will use the old Tuxedo suite
Steps will be very similar to the new one

82. Outline Transcriptomic analysis methods and tools
Transcriptome Analysis; aspects common to both assembly and differential gene expression
Download data
Quality check
Data alignment
Assembly
Differential Gene Expression
Choosing a method, the considerations…
Final thoughts and observations

83. Final thoughts and stray observations
Think carefully about what your experimental goals are before designing your experiment and choosing your bioinformatics tools

84. Final thoughts and stray observations
Think carefully about what your experimental goals are before designing your experiment and choosing your bioinformatics tools
When in doubt “Google it” and ask questions.
- Biostar (Bioinformatics explained)
- SEQanswers (the next generation sequencing community)

85. Final thoughts and stray observations
Think carefully about what your experimental goals are before designing your experiment and choosing your bioinformatics tools
When in doubt “Google it” and ask questions.
- Biostar (Bioinformatics explained)
- SEQanswers (the next generation sequencing community)
Another good resource if you are not ready to use the command line routinely is Galaxy. It is a web-based bioinformatics portal that can be locally installed, if you have the necessary computational infrastructure.

86. Final thoughts and stray observations
Today we covered how to deal with Illumina data, but you may also encounter long-read data as well
Hybrid transcriptome assemblies can be done, but are usually challenging

87. Final thoughts and stray observations
Today we covered how to deal with Illumina data, but you may also encounter long-read data as well
Hybrid assemblies can be done, but are usually challenging
R is an excellent language to learn, if you are interested in performing in- depth statistical analyses for differential gene expression analysis
Not within the scope of this lecture/lab section
We do offer a long RNA-Seq workshop that covers the “HPCBio” RNA-Seq pipeline:

88. Documentation and Support
Online resources for RNA-Seq analysis questions –
Software manuals
- Biostar (Bioinformatics explained)
- SEQanswers (the next generation sequencing community)
Most tools have a dedicated lists/forums
Thanks
Contact us at:
See website for upcoming workshops & services:

89. Thank you for your attention!
Thanks