Juicertools dump issue + How to extract normalization strategies from hic file ?

[Kristy Mualim]

Hello, 

I'm trying to process the latest hg38 Hi-C files on ENCODE and am having trouble using juicer_tools to extract the data from hic files. 

Command: java -jar juicer_tools_1.13.02.jar dump NONE ENCFF573OPJ.hic 1 1 BP 5000 chr1/chr1.observed

Traceback: Error reading datasetnull. Could not read hic file: null

Using `hic-straw`, I can confirm that the hic file is not null. It sounds like people are having multiple issues with this, as stated here: https://github.com/aidenlab/juicer/issues/57. 

Do you have suggestions to resolve this? I'm also interested in using the latest normalization strategy recommended by Erez, which is "SCALE" -- just providing this information in case it affects which version of juicer_tools I should use. 

Also, what is the command to extract the normalization strategies in the hic file? I've looked at this colab: https://colab.research.google.com/drive/1548GgZe7ndeZseaIQ1YQxnB5rMZWSsSj but it doesn't seem to have that command listed. 

Thank you! 

Best, 

Kristy

[Muhammad Shamim]

Updated the notebook with examples of dumping the normalization and expected vectors.

https://colab.research.google.com/drive/1548GgZe7ndeZseaIQ1YQxnB5rMZWSsSj

[David Katz]

Dees this have examples of dumping?

[zhang xiuxiu]

I have the same question，have you solved it?

[Moshe Olshansky]

Hi Kristy,

The correct usage is:

java -jar your-juicer_tools.jar dump observed/oe NONE ...

To explore SCALE normalisation you need a newer version of juicer_tools, say juicer_tools.2.20.00.jar