TFM_SCarbonell_Analysis

Workflow:

Processing Long-Read Data:

• ONT Basecall and QC:

  • Guppy v6 SUP was used to basecall the data with default parameters.
  • Nanoplot v.1.40.0 and MultiQC v1.13 were run using default parameters.

• Long-Read Splicing Analysis:

  • FLAIR v2.0 modules were used to process long-read sequencing data in two modes: LR-only and SR-supported.
  • Run Flair align, correct, and collapse:

    flairMODULE123.sh
    
  • Run Flair quantify:

    flairMODULE4.sh
    

• Short-Read Splicing Junctions (SJs):

  • intronProspector was used to generate SJs from short-read data using the following bash script:

    run.intronProspector.sh
    

• Data QC and Filtering:

  • SQANTI3 v5.1.2 was used to QC the long-read data models obtained by FLAIR.
  • Filter for Known (Gencode v43 annotated Genes) using SQANTI categories with the following script:

    run.sqanti.filter.known.sh
    

Processing short-read data:

Short-read sequencing data was processed using the GRAPE-NF (commit 78fd890) pipeline.

To process short-read sequencing data with GRAPE-NF, use the following command:

nextflow -bg run grape-nf -r 78fd890 --rsemSkipCi --index /nfs/users/rg/scarbonell/TFM/short_reads/metadata.tsv --genome /nfs/users/rg/projects/references/Genome/H.sapiens/GRCh38/GRCh38.p13.primary_assembly.genome.fa.gz --annotation /nfs/users/rg/projects/references/Annotation/H.sapiens/gencode43/gencode.v43.primary_assembly.annotation.gtf.gz --rg-platform ILLUMINA --rg-center-name CRG -resume -c /software/rg/grape/config/rg.singularity.dsl2.config > pipeline.log

Data Analysis

Splicing Events and Isoform Usage Analysis

• FLAIR Modules for Splicing Events and Isoform Usage:

  • FLAIR modules were employed to analyze splicing events and isoform usage.

    flairMODULE6.withSRsupport.sh
    

    flair.diff.iso.usage.KNOWN.filteredDATA.bash.sh
    

• Fisher Test for Splicing Events:

  • To conduct Fisher tests on splicing events:

    diffsplice.fishers.exact.sh
    

R Scripts for Analysis

In addition to the abovementioned tools, R scripts were utilized for further analysis and to generate data visualisations. These R scripts were crucial in providing insights and facilitating data interpretation.

Please refer to the specific R script files for detailed information on the analyses conducted and the visualizations produced.

Splicing Data Visualization

• Long-Read Data Visualization:

  • For long-read sequencing data, the FLAIR align module was run independently for each sample to obtain separate BAM files for data visualization.

    flair.align.singlebams.sh
    

• Short-Read Data Visualization:

  • Separate BAM files from short-read sequencing data were obtained directly from Grape-nf output.

• BAM File Merge:

  • Samtools was used to merge sample replicate BAM files from different compartments prior to data visualization.

• Splicing Events and Isoform Usage Visualization:

  • USCS browser and IGV were used to inspect splicing events and isoform usage. The exact coordinates for plotting were extracted and utilized to run the sushi function on ggsashimi.

    run.ggsashimi.sh
    

R-scripts:

• PCA_scaled_centered.R:

  • Purpose: Build scaled and centered PCAs for LRonly and SRsupported data.

• Rscript_different_isoform_usage.R:

  • Purpose: Perform different isoform usage analysis.

• Splicing_events.R:

  • Purpose: Analyze splicing events (IR, ES, Alt3, and Alt5).

• Upset_plots_addingGENCODE.R:

  • Purpose: Prepare upset plots of isoform counts and GENCODE for LRonly and SRsupported data.

• PLOT_regression_plots_replicates.R:

  • Purpose: Generate regression plots for LRonly and SRsupported data.

Bash-scripts:

• flairMODULE123.sh:

  • Purpose: Bash script to run FLAIR along, correct, and quantify modules from independent long-read fastq by simultaneously processing all samples and building a unique set of transcript models.

• flairMODULE4.sh:

  • Purpose: Bash script to run FLAIR quantify module, which performs isoform quantifications matrix from transcript models file.

• run.intronProspector.sh:

  • Purpose: From BAM/SAM generates a list of splicing junctions (SJs).

• sqanti3.withSRsupport.sh:

  • Purpose: Bash script to run SQANTI3 (example for short-read supported data).

• run.sqanti.filter.known.sh:

  • Purpose: Filter known genes using SQANTI categories.

• flairMODULE6.withSRsupport.sh:

  • Purpose: Bash script to run FLAIR diffSplice - module 6 (example for short-read supported data).

• flair.diff.iso.usage.KNOWN.filteredDATA.bash.sh:

  • Purpose: Bash script to run FLAIR different isoform usage tool (example for known genes filtered data).

• diffsplice.fishers.exact.sh:

  • Purpose: Compute Fisher test on splicing events.

• flair.align.singlebams.sh:

  • Purpose: From long-read fastq, generates single BAMs for each sample using FLAIR align module 1.

• run.ggsashimi.sh:

  • Purpose: Uses sushi function to generate ggsashimi plots.