RNAseq alignment and contrasts

Scripts for running nf_core pipelines in ./nf_core_scripts/ folder

Use ./makeSampleSheet.R to create samples sheet and contrast file for running nf_core pipelines.

nf_rnaseq_submit.sh runs the alignment pipeline on the cluster with STAR and salmon

nf_differentialabundance.sh runs DESeq2 for specific contrasts

gatherVolcanos.py pulls out volcano plots from nf_core subdirectories into a single pdf.

Exploratory Analysis

Scripts in ./scripts_exploratory_analysis/ folder

postProcessing.R annotates results tables with additional data from gtf (./custom/txt/*_annotated.tsv) and combines them into a single results file (./custom/rds/*.results_annotated.RDS). Also produces a text file with number of significant up/down regulated genes in each contras (./custom/txt/summaryUpDownDiffThresolds.txt)

basicExploratoryPlots.R produces:

• static and interactive volcano plots at different zooms with more info than the pipeline volcanos (./custom/plots/volcano/)
• boxplots and barplots for LFC and up/down counts by chromosome (./custom/plots/byChromosome/)
• boxplots and barplots for LFC and up/down counts by chromosome region type (tip, arm, center) (./custom/plots/byChrRegion/)
• heatmaps of significantly changing genes (./custom/plots/heatmaps/)

averageBigwigs.sh produces:

• average of bigwigs created by STAR for forward strand from all replicates (note: STAR labelling of forward and revese is backwards) (./custom/tracks/bigwigs_fr/*.forward.avr.bw)
• average of bigwigs created by STAR for reverse strand from all replicates (./custom/tracks/bigwigs_fr/*.reverse.avr.bw)
• avereage of average forward and average reverse strand bigwigs to create single track for each sample (./custom/tracks/bigwigs_fr/*.average_fr.bw)
• log2 fold change bigwigs of the different samples (./custom/tracks/bigwigs_compare/*.average_fr.log2fc.bw)

datasetComparisons.sh produces:

• correlation plots for pairs of samples (./custom/plots/correlation)
• upset plots for all N2 contrasts and lin-61 contrasts (./custom/plots/upset/)
• venn diagrams for up/down regulated genes for lin-61 contrasts (4-way) (./custom/plots/venn/)
• euler diagrams for triplets of up/down regulated genes fir kub-61 contrasts (all possible triplets) (./custom )

exploreChIP.R produces:

• bedfiles of singificantly up/down regulated genes from each mutant to be used by deeptools (./custom/exploreChIP/bed/)
• boxplots of LFC of genes from different datasets that overlap stratified 10kb triple peak regions, plotted either for all genes or only significantly upregulated autosomal arm genes (./custom/exploreChIP/*triplPeaks10kb_boxplot.pdf)
• histograms of the number of peaks per gene from each of the three peak-bedfiles plotted for all genes (./custom/exploreChIP/allGenes_bedPeakOverlap_histogram.pdf)
• stacked barplots showing the number of differnt types of peaks overlapped per gene for triple H3K9me2/hpl2/lin61 combinations or just H3K9me2/hpl2 combinations, plotted for all genes or only significantly up/down regulated autosomal arm genes. (./custom/exploreChIP/bedPeakTypeOverlap_barplot.pdf)

deeptools_sigGenes_v_ChIP.sh produces:

• deeptools heatmaps+profiles for ChIPseq of all histone marks over gene bodies +- 1kb for significantly up/down regulated genes in all four mutant contrasts. (./custom/exploreChIP/*publicChIPseq_heatmap_sort*.pdf)

Final figures

Scripts in ./scripts_finalFigures/ folder

Focussing on significant genes (padj<0.05, LFC>0.5) on autosomal chromosome arms.

Using only 4 contrasts:

Only Dimerisation, Only LLPS, No LLPS No Dimerization and Double mutant compared to lin-61; hpl-2::gfp control.

upregulatedGenesOnArms.R produces:

• Heatmap for all genes significantly upregulated in at least one group (combined or split by group)
• Boxplots of LFC of upregulated genes (or all expressed autosomal arm genes)
• Upset plot for all datasets
• Venn diagram for all datasets
• Pairwise Euler diagrams for all datasets
• Gene lists for significant genes in each group for GO and GSEA

output is all in ./custom/finalFigures/upregulatedOnArms/

integrationWithChIP.R produces:

• euler plot for overlap between genes with peaks in different ChIP seq datasets (lfcByGroup_autosomalArmTripleChIPpeaks_padj0.05_lfc0.5.pdf)
• boxplot of LFC expression for core triple-peak genes in each mutant RNAseq (lfcByGroup_autosomalArmTripleChIPpeaks_padj0.05_lfc0.5.pdf)
• same as above but with ChIP seq signal stratified into three quantiles (Q1 lowwest-Q3 highest) (lfcByGroup_autosomalArmTripleChIPpeaks_padj0.05_lfc0.5_quantile.pdf)

output is all in ./custom/finalFigures/integrationWithChIP/