filter.dccm() clarification

[Former user]

Greetings,

I've been comparing cross-correlation plots after using filter.dccm() with and without a contact map. I notice that including a contact map (with every other option the same) actually results in more non-zero correlations on the plot. This makes me think I'm misunderstanding the purpose of including a contact map at this step. I thought that the contact criteria filter out certain long-range correlations between residues that are not actually in contact. Shouldn't filtering with a contact map result in equal or fewer non-zero correlations than filtering without the contact map?

[Xinqiu Yao]

Hi,

If you have only one 'DCCM' to filter, then yes, it should have fewer non-zero values with contact map. But, if you have multiple 'DCCMs', the calculation will be very different between with and without contact map. In some cases, using a cmap may result in more non-zero values. See help(filter.dccm) for more details.

Let me know which case you are. Also, it would be helpful if you could upload an example for us to reproduce your problem.

[Xinqiu Yao]

Not a bug. Closed because the use has no response