problems with stacking parameters of a MD-history

[Miguel]

We're using X3DNA to analyze a MD-history of a DNA molecule.

When analyzing the data, we've noticed that for some frames, X3DNA does not
print out parameters. We're specially interested in the parameters that describe
stacked base-pairs (e.g. rise, twist, etc.)

We suspect that X3DNA does not print out parameters such as rise when the
corresponding frame contains stacked base-pairs that are too far from each other.
This is similar to the problem mentioned in Q6 of FAQ, which is related to
H-bonding. Q6 is solved by changing the H-bonding criterion in the file misc_3dna.par

Is it possible to change the default value of rise  so as to allow measuring the
rise of stacked bp that are further than usual? (Is it possible to do something similar for
tilt, role, twist, etc.)

Thanks
Miguel

[xiangjun]

Hi Miguel,

As suggested in the forum welcome message, please provide a (minimum) reproducible example so others can understand exactly what you mean in order to help you out.

Xiang-Jun

[Miguel]

Xiang-Jun,

thanks.

We are analyzing a short ds DNA segment from MD simulations. The
sequence is this:
A1-T10
A2-T9
A3-T8
A4-T7
A5-T6

For some frames, the base pairs in the two ends (i.e. A1-T10, or A5-T6)
are broken and find_pair cannot find them. Therefore only information
for A2-T9, A3-T8, and A4-T7 are output. For example, the rise between
A1-T10 and A2-T9 is not printed out.

We modified some of the max allowed values in the parameters in misc_3dna.par
and redo the analysis. For example, we changed the H-bonding criterion. We have
found that if the criterion value is increased just a little, find_pair still
can not find the broken basepairs in the end. But if it's increased too
much, find_pair will find some unreasonable pairs, such as A5-T7 etc.

The problem above could be solved by finding out the "optimized H-bonding
criterion" for this particular frame, so reasonable base-pairs are found. However,
this optimized value is likely to change when one is analyzing another frame.

Thanks
Miguel

[xiangjun]

Hi Miguel,

Now I understand the problem you face. If you search the forum, you will find a similar thread on this issue.

Basically, since all the MD trajectories correspond to the same base-pairing patterns (just as the different NMR models in an ensemble), you do not need to run "find_pair" for each structure. In fact, you could simply use the same paring information for each structure by just changing the first two lines in what would be the "find_pair" output. In other words, prepare an input file to "analyze" directly -- as if you do not have access to "find_pair".

HTH,

Xiang-Jun