uneuploidy_quaich_config.yml

annotations.tsv

samples.tsv

samples_annotations.tsv

uneuploidy_quaich_config.yml

# Setup the genome and view to use for analyses
genome: hg38
path_genome_folder: resources/genome/
path_genome_fasta: ../../common_data/genomes/hg38/hg38.fa
chromsizes: resources/genome/hg38_chromsizes.txt
view: resources/hg38_arms.txt

# Samples file with two columns: sample name ("sample") and path/url with cool file ("file").
# If it's URL, it will be downloaded. If the file is already downloaded it will not be overwritten.
# The downloaded file will be stored as {coolers_folder}/{sample}.mcool
samples: config/samples.tsv
# Setting up pairwise comparisons of samples
# We just have two samples, so no need to ensure certain fields match, so we set this
# setting to null
fields_to_match:
    - celltype
    - sex
# To compare by cell type, we list the name of the column with cell type from the
# samples.tsv file, and indicate which cell type we consider as reference for the
# comparisons
fields_to_differ:
    chromosomes: euploid # reference value us used to set "control" samples, with the rest being "treatment"
# Annotations file with two columns: annotation name ("bedname") and "file" (URLs or local file)
# Downloaded bed files will be stored in beds_folder
annotations: config/annotations.tsv
# Rules about correpondence between samples and annotations can be specified here
samples_annotations_combinations: config/samples_annotations.tsv

# folder definition is optional
# by default everything is done using inputs in inputs/
# and results saved in results/
inputs_folder: inputs
beds_folder: inputs/beds
bedpes_folder: inputs/bedpes
coolers_folder: inputs/coolers

project_folder: results

expected_folder: results/expected
pileups_folder: results/pileups

eigenvectors_folder: results/eigenvectors
eig_profiles_folder: results/eigenvectors/eig_profiles
compartments_folder: results/eigenvectors/compartments
saddles_folder: results/eigenvectors/saddles
pentads_folder: results/eigenvectors/pentads

insulation_folder: results/insulation
tads_folder: results/insulation/tads
boundaries_folder: results/insulation/boundaries

dots_folder: results/dots

expected:
    do: True #If False, still will calculate them when needed by other rules
    cis: True
    trans: True
    resolutions:
        - 1000
        - 10000
        - 100000
        - 250000
        - 500000
        - 1000000
    extra_args_cis: "--smooth --aggregate-smoothed --clr-weight-name weight_cis"
    extra_args_trans: "--clr-weight-name weight_trans"

# Parameters for calculation of the eigenvectors (cooltools eigs-cis or eigs-trans)
# They are phased using GC content whcih is automatically calculated from the genome .fastqs
eigenvectors:
    do: True
    cis: True
    trans: False
    resolutions:
        - 100000
        - 250000
        - 500000
        - 1000000
    save_compartment_beds: True # If True, will run a simple HMM to save a .bed file
                                # with coordinates of A and B compartments

# Parameters for saddles - a way to quantify global compartment structure (cooltools saddle)
saddles:
    do: True
    bins:
        - 100
    distance_limits:
        - 1600000
        - 102400000
    range: "--qrange 0.01 0.99 "
    extra: "--strength"

# Parameters for pentads, another way to average compartments
# Described in https://doi.org/10.1186/s12859-022-04654-6, but reimplemented using coolpuppy here
pentads:
    do: False
    do_diff: False
    data_resolution: 100000
    eigenvector_resolution: 250000
    groupby:
        - chrom1
    norms:
        - expected
    modes:
        - local
        - distal

# Parameters for pileups using coolpuppy
pileups:
    do: True
    resolutions:
        # - 10000
        - 50000
        # - 100000
    distance_limits:
        - 500000
        - 100000000
        # - 12500
        # - 102400000
    shifts: 0
    expected: True
    ooe: True #Only applies when expected is True
    arguments: # Different modes of pileups that can be applied to different annotation
               # .bed files, relationship defined in samples_annotations_combinations file above
        # distal: "--mindist 500000 --clr-weight-name weight_cis"
        # by_distance: "--by_distance --clr-weight-name weight_cis"
        # local: "--local --clr-weight-name weight_cis"
        local_rescaled: '--local --rescale --rescale_pad 1 --clr-weight-name weight_cis'
        local_rescaled_by_chrom_group: "--local --rescale --rescale_pad 1 --clr-weight-name weight_cis --groupby chrom_group1"

        distal_rescaled: '--mindist 500000 --rescale --rescale_pad 1 --clr-weight-name weight_cis'
        distal_rescaled_by_chrom_group: '--mindist 500000 --rescale --rescale_pad 1 --clr-weight-name weight_cis  --groupby chrom_group1'
        # by_strand_by_distance: "--by_strand --by_distance"
        # by_strand_local: "--by_strand --local"
        # by_strand_distal: "--by_strand --maxdist 1000000"
        # by_window_short_range: "--by_window --subset 1000 --maxdist 2000000"
        # by_window_long_range:  "--by_window --subset 1000 --mindist 2000000"

# Parameters for calling dots/loops, using different methods
dots:
    methods:
        cooltools:
            do: False
            extra: "--max-loci-separation 10000000 --fdr 0.02"
        chromosight:
            do: False
            extra: ""
        mustache:
            do: False
            max_dist: 10000000
            extra: "-pt 0.05 -st 0.8"
    resolutions:
        - 10000
    pileup: True


# Parameters to calculate insulation score and how to threshold (cooltools insulation)
insulation:
    do: False
    resolutions:
        1000:
            - 3000 # What window sizes to do for what resolution
            - 5000
            - 10000
            - 25000
        10000:
            - 50000
            - 100000
    extra: '--chunksize 1000000000 --threshold Li --ignore-diags 2' # For the threshold,
                                                                    # choose between a specific float, Li or Otsu,
                                                                    # see https://github.com/open2c/open2c_examples/blob/master/Insulation_and_boundaries.ipynb
    pileup: True # Whether to pileup thresholded boundaries


# Find differential boundaries between pairs of samples using a simple presence/absence or FC threshold for boundary strength
compare_boundaries:
    do: False
    fold_change_threshold: 5
    pileup: True

# Combine neighbouring strong insulation score boundaries into TADs
TADs: 
    do: False
    resolutions:
        10000: # What window sizes to do for what resolution
            - 50000
            - 100000
            - 1000000
    max_tad_length: 1500000
    pileup: True