Thanks for your kind words!
For any kind of dense file, calling loops via input bedpe is going to be pretty space intensive. But you can try by converting the merged_nodups into bedpe. It should be fairly straightforward.
The format is here:
I assume for the
bedpe files, you would take input fields 2 and 6 (for chromosome) and convert to output fields 1 and 4; and then output fields 2/3 would be the position (field 3) plus whatever resolution (you could do just +1, or + read length), and output fields 5/6 would be the same but from the second position (field 7). Strand1 and strand2 (output fields 9/10) are given by input fields 1 and 5.
Another possibility is to take binned data from hic via
Juicer Tools Dump or
Straw and convert that to your bedpe, in which case you would use the beginning and end of bin number * resolution and the count would be the score.
Hope that helps!
Best
Neva