This can also be used for Hi-ChIP experiments, in that case the output with matrix with iced values is ignored and the matrix to use is matrix with raw values .

Input datasets

• The workflow needs a list of dataset pairs of fastqsanger.

Input values

• genome name: suggested from the bowtie2 indices, it is used to map and build the list of bins.

• restriction enzyme: Restriction enzyme used e.g. A^GATCT,BglII. The '^' is used to express where the enzyme cuts.

• No fill-in: If you used a biotin fill-in protocol, put this to false, else, put it to true.

• minimum MAPQ: Filtering to apply to pairs you want to keep in your matrix, set it to 0 to not apply filtering (HiCUP already filter for uniquely mapped or MAPQ30).

• Bin size in bp: Used to generate your first matrix but you will be able to rerun the subworkflow hic_tabix_to_cool_cooler to get other resolutions.

• Interactions to consider to calculate weights in normalization step: this is a parameter for the last correction step (ICE).

For the region capture workflow:

• chromosome, start and end positions of the capture region

For the Hi-C workflow:

• region to use in pyGenomeTracks to check the matrices.

Processing

• Reads are processed with HiCUP which comprises these steps:

  • Truncation of reads for the religation motif

  • Mapping of mates independently with bowtie2

  • Pairing the mates when both mates are uniquely mapped or MAPQ30

  • Filtering the pairs for undigested, self-ligated...

  • Removing duplicates

• The output BAM file is converted to medium juicer format: <readname> <str1> <chr1> <pos1> <frag1> <str2> <chr2> <pos2> <frag2> <mapq1> <mapq2> where str = strand (0 for forward, anything else for reverse) and pos is the middle of the fragment.

• The pairs are filtered for MAPQ if specified.

• For the region capture Hi-C workflow the pairs are filtered for both mates in the captured region.

• The filtered pairs are sorted and indexed with cooler_csort.

• The pairs are loaded into a matrix of the given resolution and balanced with cooler.

• A final plot is made with pyGenomeTracks using the balanced matrices on the region provided or the capture region.

Subworkflows

There are 2 subworkflows: hic_tabix_to_cool_cooler and hic_fastq_to_pairs_hicup.ga .

hic_tabix_to_cool_cooler

This first subworkflow can be used to generate matrices to different resolutions using one of the output of the full workflow ( valid pairs filtered and sorted ).

If the dataset are still in galaxy (format: juicer_medium_tabix.gz), the workflow can be run directly.

If the dataset is not anymore in galaxy, you need to upload and specify the datatype as: juicer_medium_tabix.gz

hic_fastq_to_pairs_hicup

The second subworkflow has no real reason to be launched by itself except for QC tests.

If you want to run the first subworkflow from these results:

• You first need to filter the pairs ( valid pairs in juicebox format MAPQ filtered ) for the capture region if relevent using the tool Filter1 ( Filter data on any column using simple expressions) with the condition (c3=='chr2' and c4<180000000 and c4>170000000) and (c7=="chr2" and c8<180000000 and c8>170000000) if your capture region is chr2:170000000-180000000.

• Then you need to run cooler_csort ( cooler csort with tabix Sort and index a contact list.) with as input the valid pairs in juicebox format MAPQ filtered or the output of the previous step and for "Format of your input file" use "Juicer Medium Format".

The output of cooler_csort can be used as input of the first subworkflow.