DNase Hi-C

[kalavattam]

Hi, thank you for the great program.

Do you have any recommendations, advice, or considerations for using mHi-C with Hi-C datasets generated with the in situ DNase Hi-C protocol (which uses DNase I instead of restriction enzymes)?

[yezhengSTAT]

Hello,
That is a great question. The mHi-C model should apply but the pipeline will need to be adjusted for DNase Hi-C protocol. Currently, mHiC uses the restriction enzyme cutting site to check if the read alignment position is valid (i.e., if the candidate alignment position is too far away from the enzyme cutting site) and the enzyme recognization pattern is used for rescuing chimeric reads (this function can be turned off). Therefore, you will probably need to customize the third step of the pipeline to make it run on DNase Hi-C data.

Hope it is helpful. ;)

Thanks,
Ye

[dborgesr]

Just wanted to chime in and say i ran into the same problem while trying to run Micro-C data through the pipeline and see if you were able to get it running?
If i get some free time i'll try to edit step three to allow no fragment files (need to edit the input parameter checking as well as the main loop to not check for fragment lenghts etc)

[lucidif]

Hello, I am considering using your workflow for microC data analysis. I was thinking of skipping steps 1, 2, and 3 (using instead a custom pipeline for microC data analysis). I would like to start from step 4, thus also skipping all the filters related to restriction fragments. The issue is that even in step 5, fragment files are required (for example, in the case of the generate_FragPairs process). Do you think it would be correct if I pass bin coordinates instead of fragment files? In this way, each bin would correspond to a "fake" restriction fragment. This could be a solution for me to start seeing how your model works with my data (unless this trick affects the model's functioning). Finally, how could I completely remove the processes that require fragfile in step 5? Thank you very much!

[yezhengSTAT]

I do not see obvious drawback of defining the read alignment coordinate at the bin level instead of basepair level. You probably can define a binFile in step 5 by aggregating the fragments falling within the same bin.

Thanks,
Ye