Raw matrix from TADbit API and TADbit tools are different.

[pollicipes]

Hi again, David:

This might (probably) be related to this previous issue #377 (comment)

When I get raw matrices for a same chromosome using the TADbit Python API and using TADbit tools bin the matrices that I get are quite different. You can also see that the matrix obtained with the TADbit API is more dense.

MATRIX FROM TADBIT TOOLS:

MATRIX FROM API:

You can also see the first 10x10 entries for both matrices: Some cells contain the same values, but others just don't. Some are different by orders of magnitude:

Thank you! :)

[david-castillo]

Hi Juan,

My bet is that it comes from the same issue. Can you paste the code that you use for the api so I can confirm it's the same problem?

Thanks

David

[pollicipes]

Hi David!

Yes, this is basically it:
I just load a bam containing only the valid pairs on the bam variable.

hic_data1 = load_hic_data_from_bam(bam,
                                   resolution=reso,
                                   region=region,
                                   ncpus=6)

hic_map(hic_data1,
        normalized=False, 
        resolution=reso,
        savedata="outfile.mat")