Restriction map and bait map - Hi-C

Question

[Kevin Blighe]

I have been running my first Hi-C analysis recently and it seems that I am unsure on one specific term, for which I would appreciate some clarification.

For a program that I am using, CHiCAGO ( https://bioconductor.org/packages/release/bioc/html/Chicago.html ), a restriction map and bait map is required. I have the restriction map. The DOCs say that the bait map is a subset of the restriction map. Thus, I assume that the bait map is just where the contact points have been determined, so, it would consist of overlapping the aligned BAM reads with the restriction map?

Obrigado, Kevin

Answer 1

[ATpoint]

The baits in the promoter-capture HiC context are the affinity beads one uses to pull out the promoter-associated fragments from the HiC library, therefore the name capture HiC. That way one enriches for those interactions that involve gene promoters, effectively reducing the super complex HiC libraries to a few percent of total. The bait map therefore consists of all those intervals from the restriction map that overlap with the coordinates of your baits. The bait coordinates must be provided by the person who designed them. I think in the past I had a file with the sequences of the baits and then simply mapped them to the genome build I used with bowtie2. Since they are designed against the reference genome and quite long you should get 100% identify matches for each of them.

There is a script within chicagoTools (link) called create_baitmap_rmap.pl that automates the baitmap creation if you provide a (I think) BED file with the bait coordinates and the output of the genome in silico digestion produced by HiCUP for the enzyme that was used in the experiment.

[Kevin Blighe]

Yes, I was using the create_baitmap_rmap.pl script but it requests an 'oligos' file, in addition the the HiCUP digest file. Unfortunately, I cannot seem to find the oligos file anywhere, relating to this study: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE86821

I think that i will have to follow-up on this point:

For PCHi-C, ligation products containing promoter sequences were captured using a custom-made RNA capture bait system (SureSelect Target Enrichment, Agilent Technologies) designed as previously described in Mifsud, B. et al. Nat. Genet. 47, 598–606 (2015), and containing 37,608 biotinylated RNAs targeting 22,076 HindIII restriction fragment ends.

[ATpoint]

People always use (from what I know) the baits that Mifsud and colleagues used in their original work. You probably want the human coordinates in Suppl. Table 4 from this study here. It contains sequences and coordinates. I think this was hg19, that is why I aligned the sequences to hg38. I felt safer with realigning rather than liftover. As said the sequences are long and unique, so you should get perfect matches for all of them.

Supplementary Table 4 => Coordinates and sequences of SureSelect baits used to capture promoter fragments.

[Kevin Blighe]

Du bist ein genie

[ATpoint]

You're welcome. What I just remembered is that you must not have bedtools v.2.26 installed (or in PATH). There was any issue with it that the authors mention somewhere in the manual. Be sure to have 2.25 or earlier.