search

XiaoTaoWang / EagleC

A deep-learning framework for predicting a full range of structural variations from bulk and single-cell contact maps
Other
51 stars 8 forks source link

Hi-C 3.0 with DdeI+DpnII #34

Open gibcus opened 5 months ago

gibcus commented 5 months ago

Xiaotao, This looks like an amazing tool and I would love to give a shot on our newer Hi-C protocol that uses DdeI and DpnII. I know your code can work with Arima, so is there something you or I can do to make this work for Hi-C with DdeI and DpnII dual digestion?

Johan

XiaoTaoWang commented 5 months ago

Hi Johan,

Since the contact patterns used by EagleC for SV detection are not likely to be affected by the enzymes used in Hi-C, I believe the current models can be directly applied to detect SVs on your matrix. Please give it a try and let me know if you have any problems in running the program :)

Xiaotao

gibcus commented 5 months ago

It seems to work if I use ICE! However, if I want to use CNV, I'd have to use neoloop finder and specify an enzyme, right? Using neoloop finder would be great, because to do a good compartment analysis on some of my coolers, I would need to remove the CNV bias.

XiaoTaoWang commented 5 months ago

I see. Yes, you need to run NeoLoopFinder to perform CNV normalization. I will update it soon to make it work for "DdeI+DpnII" and keep you informed.

Xiaotao

gibcus commented 4 months ago

Hi Xiaotao, I'm folllwing up on the above. Please let me know if it would work if I adapt calculate-cnv by adding Dde-DpnII to this line:

parser.add_argument('-e', '--enzyme', default='MboI', choices = ['HindIII', 'MboI', 'DpnII', 'Arima', 'BglII', 'uniform'],
                        help='''The restriction enzyme name used in the Hi-C experiment. If the genome
                        was cutted using a sequence-independent/uniform-cutting enzyme, please consider
                        setting this parameter to "uniform".''')

And provide a weblink with 'hg38.DpnII-DdeI.npz'

XiaoTaoWang commented 4 months ago

Hi Johan,

Sorry for the delay. You were right; that's all we need to make it work for Dde-DpnII. I have generated the npz file, which you can download from the link https://www.jianguoyun.com/p/DS6GgfsQpPuuDBjNlsIFIAA. However, I haven't had the time to test it. Please feel free to download the file and modify the script as needed.

Xiaotao