paper repository for Vazquez-Garcia & Obermayer et al.

contents:

• data: contains input data for paper figures (prepared using prepare_data.R

• paper_figures.Rmd: R code to produce paper figures (uses data in data, nothing else needed)

• swibrid_runs: contains config files for various SWIBRID runs on human or mouse data, or the simulations

  • benchmarks: config files for the benchmarks

    • dense: using dense MSA, can be run as is using swibrid test in that folder
    • sparse: using sparse MSA. for this, the sparsecluster package needs to be installed

  • mouse: config files for mouse data

    • download raw fastq files from SRA (accession PRJNA1190672) into raw_data and run demultiplex_dataset.sh; this will put fastq and info.csv files for individual samples into input and make it possible to run all samples in one go
    • download mm10 genome from UCSC or elsewhere
    • download gencode M12 reference and use swibrid prepare_annotation
    • use config.yaml for running all mouse data (assumed to produce the folder output)
    • use config_noSg.yaml for running everything only on Sm + Sa (potentially restrict info files in input to reads with Sa primer; assumed to produce the folder output_no_Sg)

  • human: config files for human data and various scripts to create plots

    raw sequencing data for human donors cannot be shared due to patient privacy legislation

    • demultiplex_dataset.sh is used to demultiplex input for each run, demultiplexed fastq and info.csv files would be expected in input
    • get hg38 genome and gencode v33 reference, create LAST index
    • config.yaml for "regular" runs (assumed to produce the folder output)
    • config_reads_averaging.yaml to use averaging of features over reads not clusters (assumed to produce the folder output_reads_averaging)
    • combine_replicates.sh to pool reads from technical replicates
    • plot_bars.sh and plot_bars.py to plot isotype fractions as in Fig. 1
    • plot_circles.sh and plot_circles.py to create bubble plots of Fig. 1
    • plot_clustering.sh to create read plots for Fig. 1 and S2
    • plot_breakpoints.sh and plot_breakpoint_stats.py to create breakpoint matrix plot of Fig. 2A
    • meta_clustering.py and meta_clustering.sh is used for meta-clustering of clusters from multiple samples
    • cluster_tracing.ipynb is used to create the annotated read plots for tracing clusters across samples
    • minION_vs_pacBIO_vs_HTGTS.ipynb is used to compare different technologies in different species and regions
    • external: config files for public datasets (Vincendeau et al. and Panchakshari et al.)
    • for Vincendeau et al., download data from SRA (PRJNA831666) into the Vincendeau subfolder and run make_info.py on every sample to create dummy files with primer locations
    • for Panchakshari et al., use get_data.sh in the HTGTS folder to download data (accessions SRR2104731-47 and SRR6293456-63), collapse read mates with bbmerge and create info files

• supplementary_note.ipynb: python code to make plots for supplementary note (needs numpy, scipy, pandas, seaborn)