Hi,
I've tried SNAP on
1ffk and did not find any H-bonds between proteins and RNA, as you reported. However, trying SNAP on
1s72 reports 1749 H-bonds. So I guess it must be due to atomic coordinates issue with 1ffk. You may want to dig into 1ffk deeper to understand why, but I believe SNAP is doing its job as designed.
Hope this help.
Xiang-Jun
Did you know that you could also use DSSR with JSON output to easily parse H-bonds?
x3dna-dssr -i=1oct.pdb --json | jq '.hbonds[] | select(.residue_pair == "nt:aa")'
The underlying algorithm for H-bond detection is the same for DSSR and SNAP, with the default text output tailored to nt:nt and nt:aa respectively. With the DSSR
--json option, 'ALL' H-bonds are reported. In the near future, the
--json will be available to SNAP as well.