coherence function of the filtered signal

Dear All, I am just wondering if calculation coherence between two signal after loww-pass filtering (maybe any filtering) does make any sense? I have 2 biomedical signal and coherence between raw signal is very low (0.13) whilst when I filter signals with lowpass with cutoff of 0.15Hz(im interested in very low frequencies) I get much higher coherence(0.7-0.8). My worry is that during filtering I remove most of the spectrum and I think that may rise the mean coherence, as removed frequencies will overlap one another. Please correct me if im wrong. Is there any way in matlab to control over which frequencies the coherence is calculated? I used mscohere for calculating coherence. Thanks in advance for comments.

[Wayne King]

are you saying that you see a coherence value in a specific frequency region that is 0.13, but then after lowpass filtering that coherence jumps to 0.7 or 0.8?

[Katarzyna]

I dont know how to calculate coherence for specific frequencies... I used mean from mscohere vector. That would be very helpful if you could explain how to specify frequencies of interest in this function. Thanks

[Jakub]

to calculate coherence for certain frequencies, use wcoh function a then recalculate scales to pseudofrequencies(scal2freq)