Galaxy RNA-Seq Analysis: Tuxedo Protocol
17 likes22,613 views
This document provides an overview of analyzing RNA-Seq data using the Tuxedo protocol in Galaxy. It describes experimental design considerations, quality control of sequencing data using FastQC, mapping reads to a reference genome using Tophat, determining differential expression with Cuffdiff, and visualizing results using IGV and CummeRbund. The tutorial walks through an example analysis on Drosophila melanogaster RNA-Seq data, covering topics such as setting file formats, running alignment and expression tools, extracting workflows, and useful Galaxy resources.
Galaxy RNA-Seq Analysis: Tuxedo Protocol
- 1. Galaxy RNA-Seq Analysis: Tuxedo Protocol ChangBum Hong, KT Bioinformatics, GenomeCloud SCIC genome-cloud.com This work is licensed under the Creative Commons Attribution-NonCommercialShareAlike 3.0 New Zealand License. To view a copy of this license, visit http:// creativecommons.org/licenses/by-nc-sa/3.0/nz/ or send a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA.
- 2. Introduction • RNA-Seq • • • Transcriptome assembly • Qualitative identification of expressed sequence Differential expression analysis • Quantitative measurement of transcript expression RNA-Seq Applications • Annotation: Identify novel genes, transcripts, exons, splicing events, ncRNAs • • Detecting RNA editing and SNPs Measurements: RNA quantification and differential gene expression
- 3. Experimental design • What are my goals? • Transcriptome assembly? • Differential expression analysis? • Identify rare transcripts? • What are the characteristics of my system? • Large, complex genome? • Introns and high degree of alternative splicing? • No reference genome or transcriptome?
- 5. Sequencing • Platforms • Library preparation • Multiplexing • Sequence reads • File names • Fastq format(Formats vary) • 4 lines per read Illumina Read ID
- 6. Data Quality Control • Data Quality Assessment • Identify poor/bad sample • Identify contaminates • Trimming: remove bad bases from read • Filtering: remove bad reads from library
- 7. Read Mapping • Alignment algorithm must be • fast • able to handle SNPs, indels, and sequencing errors • allow for introns for reference genome alignment • Input • fastq read library • reference genome index • insert size mean and stddev(for paired-end libraries) • Output • SAM (text) / BAM (binary) alignment files
- 8. Differential Expression • Cuffdiff (Cufflinks package) • Pairwise comparisons • Differnetial gene, transcript, and primary transcript expression • Easy to use, well documented • Input: transcriptome, SAM/BAM read alignments
- 9. Transcriptome Assembly • RNA-Seq • Reference genome • Reference transcriptome • RNA-Seq • Reference genome • No reference transcriptome • RNA-Seq • No reference genome • No reference transcriptome
- 10. Reference Genome FASTA GFF/GTF Experimental Design Referecne Transcriptome RNA Sequencing FASTQ Reads FASTQ Data Quality Control Tuxedo protocol
- 11. Combining tools in a pipeline • Linux Command-line Tools • Shell script, Makefile • GUI Based pipeline • DNANexus • SevenBridegs Genomics • Galaxy • Open Source • Wrapper for command line utilites • Workflows • Save all steps you did in your analysis • Return the entire analysis on a new dataset • Share your workflow with other people
- 12. How to use Galaxy? GALAXY MAIN: User disk quotas 250GB for registered users, maximum concurrent jobs: 8 NO WAIT TIMES NO NO JOB STORAGE SUBMISSION QUOTAS LIMITS NO DATA TRANSFER BOTTLENECKS NO IT EXPERIENCE REQUIRED NO REQUIRED INFRASTRUCTURE COST GALAXY MAIN Free LOCAL GALAXY Free ? CLOUD GALAXY (AMAZON) 동일사양 대비 약 2배 (KT의) SLIPSTREAM GALAXY $19,995 (2천2백만원) KT GenomeCloud GALAXY 시간당 740원 부터
- 13. Outline of tutorial • Starting Galaxy • Mapping with Tophat • Workflows • Visualizing alignment with IGV • Computing differential expression with cuffdiff • Cuffdiff visuaalization with CummeRbund
- 15. Starting Galaxy • Tutorial Dataset • Accessing Galaxy • Import files for one sample into current history • Set file attributes • Run FastQC
- 16. Tutorial Dataset • FASTQ files (fastq): Sequence Reads • Reference (fasta): Genome Sequence (galaxy default) • Geneset (GTF / GFF3): Reference Geneset • Bowtie2 index: Reference index files for Bowtie2 (galaxy default)
- 17. Tutorial Dataset Reference & Gene sets • Ensembl • http://www.ensembl.org/info/data/ftp/index.html
- 18. Tutorial Dataset Reference & Gene sets •illumina iGenomes • The iGenomes are a collection of reference sequences and annotation files for commonly analyzed organisms. The files have been downloaded from Ensembl, NCBI, or UCSC, and chromosome names have been changed to be simple and consistent with their download source. Each iGenome is available as a compressed file that contains sequences and annotation files for a single genomic build of an organism. • http://support.illumina.com/sequencing/sequencing_software/igenome.ilmn
- 19. Tutorial Dataset Sequencing data •Sequencing data (Drosophila melanogaster) • Gene Expression Omnibus at accession GSE32038 • http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE32038
- 20. Biological replicates vs. technical replicates Technical Replicates Biological Replicates
- 21. Accessing Galaxy • • Open a web browser and navigate to Galaxy website usegalaxy.org or www.genome-cloud.com Log in with username and password select galaxy service GenomeCloud (genome-cloud.com)
- 22. when your galaxy is ready you will recive the e-mail access the galaxy via public ip address you can register via user menu > register Center pane Tools pane History pane
- 23. Import files • • Open a web browser and navigate to Galaxy website usegalaxy.org or www.genome-cloud.com Log in with username and password example fastq and gtf files are located in shared data > RNA-Seq with Drosophila melanogaster import data into your history panel (read to analysis)
- 24. Set file attributes • • In the history pane click on the pencil icon Enter “fastqsanger” (It will takes time) Sanger Phread+33 fastqsanger (cassava 1.8 ▲ ) Ilumina 1.3 Phread+64 fastqillunina (cassava 1.8 ▼) Solexa Solexa+64 fastqsolexa Tophat options --solexa-quals: Use the Solexa scale for quality values in FASTQ files --solexa1.3-quals: Phred64/Illumina 1.3~1.5 ! BWA options -l : The input is in the Illumina 1.3+ read format (quality equals ASCII-64) ! GenomeCloud (g-Analysis)
- 25. Error probability Quality Score Encoding CASAVA 1.8.2 Quality Score (or Q-score)
- 26. Run FastQC • • Load the FastQC tool from the tool pane Set the input file (repeat this step on the C1, C2 all piar files)
- 27. wait running done error Galaxy status When fastqc has finished running, click on the eye on the FastQC output file to display
- 28. illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) illumina (bad dataset in FastQC homepage) Per base sequence quality illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) illumina (bad dataset in FastQC homepage) Per sequence quality score illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) Per base sequence content illumina (bad dataset in FastQC homepage)
- 29. illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) illumina (bad dataset in FastQC homepage) Per base GC content illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) illumina (bad dataset in FastQC homepage) Per sequence GC content illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) Per base N content illumina (bad dataset in FastQC homepage)
- 30. illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) illumina (bad dataset in FastQC homepage) Sequence Length Distribution illumina (in-house data) IonTorrent (in-house data) illumina (good dataset in FastQC homepage) Sequence Duplication Levels illumina (bad dataset in FastQC homepage)
- 32. Mapping with Tophat • Initial Tophat run • Determine insert size • Rerun Tophat with correct insert size • Review mapping statistics
- 33. Initial Tophat run • • • Use Full Tophat paramters Paired-end FASTQ files, Select reference genome, Use Own Juctions(Yes), Use Gene Annotation Model(Yes) Gene Model Anntations (use GFF file)
- 34. Determine insert size • Load the insert size tool “NGS: Picard (beta) -> Insertion size meterics”
- 35. Determine insert size • • Click “eye” icon Identify the MIN_INSERT_SIZE (198)
- 36. Rerun Tophat • • • Click any one of the Tophat2 output files in the history panne Click on the circular blue arrow icon Change the “Mean Inner Distance between Mate Pairs” (198)
- 37. Tophat Output • • unmapped.bam (BAM) • junctions.bed (BED): list BED track of junctions reported by Tophat where each junction consists of two connected BED blocks where each block is as long as the max overhang of nay read spanning juction • deletions.bed (BED): mentions the last genomic base before the deletion • insertions.bed (BED): mentions the first genomic base of deletion accepted_hits.bam (BAM): a list of read alignments in BAM/SAM format
- 38. Load files into IGV • • • • Click on the “accepted hits” file in the history pane Click on the “display with IGV web current” A file named “igv.jnlp” will be downloaded by your browser Open with text editor copy BAM file location
- 39. IGV with Housekeeping gene http://www.sabiosciences.com/rt_pcr_product/HTML/PADM-000Z.html
- 40. Load files into IGV • • Enter “Act42A” in the search box to view the reads aligning Right-click on the coverage track and select “Set Data Range” (max value to 4372) Housekeeping gene: Act42A Set max value
- 42. Keyword: regucalcin (calcium-binding protein) this gene has four isoforms
- 43. Load files into Trackster • • • Click on the “accepted hits” file in the history pane Click on the graph icon and select “Trackster” Select bam files
- 44. drag into new group move to regucalcin gene create new group ‘Add group’
- 45. set max value
- 46. Run cuffdiff • • • Load the Cuffdiff tool: “NGS:RNA Analysis->Cuffdiff ” Perform replicate analysis(Yes) Add new Group / Add new Replicate
- 47. Cuffdiff output • Genes: gene differential FPKM • Isoforms: Transcript differential FPKM • CDS: Coding sequence differential FPKM
- 48. • • View and filter cuffdiff output Differential Gene Expression (DGE) Filter out genes with significant change in expression with a log fold-change of at least 1 “C14 == ‘yes’ and abs(c10)>1” in the “With following condition” text box
- 50. • • Cuffdiff visualization with CummeRbund Load the CummeRbund tool: NGS:RNA Analysis->cummerbund Plot type: Density, check the “Replicates” box
- 51. Samples have similar density distribution(density plot) Samples cluster by expression condition (MDS / PCA plot) Samples cluster by experimental condition (Dendogram)
- 52. Volcano Differential analysis results for regucalcin Expression plot shows clear differences in the expression of regucalcin across conditions C1 and C2 (four alternative isoforms) Scatter plots highlight general similarities and specific outliers between conditions C1 and C2
- 53. Extract workflow from current history • Click on the small gear icon and select “Extract Workflow”
- 54. Edit workflow • • Click on “Workflow” at the top of the Galaxy window Move the elements of the workflow
- 55. Run workflow • • • Load a workflow by clicking on “Workflow” ath the top of the screen Click on “Run” Select the input datas
- 57. Useful galaxy sites • Public main galaxy site (user disk quotas 250GB for registered users, maximum concurrent jobs: 8) • • Test galaxy site (beta site for galaxy main instance) • • http://hongiiv.tistory.com/701 Galaxy를 이용한 SNP 분석 (Korean) • • https://wiki.galaxyproject.org/Learn Galaxy를 이용한 NGS 분석 (Korean) • • https://test.galaxyproject.org/ Galaxy screen cast and tutorials • • https://usegalaxy.org/ http://hongiiv.tistory.com/652 Galaxy를 이용한 부시맨 genome 분석 (Korean) • http://hongiiv.tistory.com/655
- 58. Acknowledgements: YoungGi Kim HanKyu Choi WanPyo Hong KangJung Kim Thank you