If *overlapping nucleotides* no contacts are computed

vreinharz commented 7 years ago

In the file: zGetNTData, the warning has overlapping nucleotides and should be used with caution is produced when the array d, the atom distances has the condition if d(min(10, length(d))) < 1 but in that case the base pairs are never computed! In particular 1S72 fails with that condition. Closing the if condition instead of the else fixed the problem as here zGetNTData.txt.

I am not sure if side effects are produced.

clzirbel commented 7 years ago

Thanks for the note. The problem seems to be that in certain PDB files, nucleotides have 3D coordinates that are literally right on top of other nucleotides. This can result in many spurious basepairs, which we are reluctant to annotate. But I have never seen this cause trouble with 1S72. Could you tell me more about the 1S72 file you are using?

On 2/8/2017 1:38 PM, Vladimir Reinharz wrote:

In the file: /zGetNTData/, the warning has overlapping nucleotides and should be used with caution is produced when the array |d|, the atom distances has the condition if d(min(10, length(d))) < 1 but in that case the base pairs are never computed! In particular 1S72 fails with that condition. Closing the |if| condition instead of the else fixed the problem as here zGetNTData.txt https://github.com/BGSU-RNA/FR3D/files/761598/zGetNTData.txt.

I am not sure if side effects are produced.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/BGSU-RNA/FR3D/issues/11, or mute the thread https://github.com/notifications/unsubscribe-auth/AA8PkzGFGC-wflPmp0q_U8qxIIFcEGY5ks5ragu7gaJpZM4L7Mo2.

--

Craig L. Zirbel Professor and Graduate Coordinator Department of Mathematics and Statistics BGSU | Bowling Green State University Bowling Green, OH 43403 +1-419-372-7466; zirbel@bgsu.edu mailto:zirbel@bgsu.edu http://www-math.bgsu.edu/~zirbel http://www-math.bgsu.edu/%7Ezirbel

vreinharz commented 7 years ago

Hi Craig,

Both the latest .cif from the PDB and the .pdb file you provide with the repositor. Thye thing is that 1S72 has already in the repository the PrecomputeData, so the error doesn't show unless you erase it first.

Best,

Vladimir Reinharz

On Feb 8, 2017, at 9:46 PM, Craig L. Zirbel notifications@github.com wrote:

Thanks for the note. The problem seems to be that in certain PDB files, nucleotides have 3D coordinates that are literally right on top of other nucleotides. This can result in many spurious basepairs, which we are reluctant to annotate. But I have never seen this cause trouble with 1S72. Could you tell me more about the 1S72 file you are using?

On 2/8/2017 1:38 PM, Vladimir Reinharz wrote:

In the file: /zGetNTData/, the warning has overlapping nucleotides and should be used with caution is produced when the array |d|, the atom distances has the condition if d(min(10, length(d))) < 1 but in that case the base pairs are never computed! In particular 1S72 fails with that condition. Closing the |if| condition instead of the else fixed the problem as here zGetNTData.txt https://github.com/BGSU-RNA/FR3D/files/761598/zGetNTData.txt.

I am not sure if side effects are produced.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/BGSU-RNA/FR3D/issues/11, or mute the thread https://github.com/notifications/unsubscribe-auth/AA8PkzGFGC-wflPmp0q_U8qxIIFcEGY5ks5ragu7gaJpZM4L7Mo2.

--

Craig L. Zirbel Professor and Graduate Coordinator Department of Mathematics and Statistics BGSU | Bowling Green State University Bowling Green, OH 43403 +1-419-372-7466; zirbel@bgsu.edu mailto:zirbel@bgsu.edu http://www-math.bgsu.edu/~zirbel http://www-math.bgsu.edu/%7Ezirbel

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/BGSU-RNA/FR3D/issues/11#issuecomment-278440194, or mute the thread https://github.com/notifications/unsubscribe-auth/AAj8xM56RwR5MYrTnDTzNIskV7RZbR_cks5rahuLgaJpZM4L7Mo2.

vreinharz commented 7 years ago

Oh, and I am using the script annotate_pdb.m from Anton https://gist.github.com/AntonPetrov/4108400 https://gist.github.com/AntonPetrov/4108400

Vladimir Reinharz

On Feb 8, 2017, at 9:46 PM, Craig L. Zirbel notifications@github.com wrote:

Thanks for the note. The problem seems to be that in certain PDB files, nucleotides have 3D coordinates that are literally right on top of other nucleotides. This can result in many spurious basepairs, which we are reluctant to annotate. But I have never seen this cause trouble with 1S72. Could you tell me more about the 1S72 file you are using?

On 2/8/2017 1:38 PM, Vladimir Reinharz wrote:

In the file: /zGetNTData/, the warning has overlapping nucleotides and should be used with caution is produced when the array |d|, the atom distances has the condition if d(min(10, length(d))) < 1 but in that case the base pairs are never computed! In particular 1S72 fails with that condition. Closing the |if| condition instead of the else fixed the problem as here zGetNTData.txt https://github.com/BGSU-RNA/FR3D/files/761598/zGetNTData.txt.

I am not sure if side effects are produced.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/BGSU-RNA/FR3D/issues/11, or mute the thread https://github.com/notifications/unsubscribe-auth/AA8PkzGFGC-wflPmp0q_U8qxIIFcEGY5ks5ragu7gaJpZM4L7Mo2.

--

Craig L. Zirbel Professor and Graduate Coordinator Department of Mathematics and Statistics BGSU | Bowling Green State University Bowling Green, OH 43403 +1-419-372-7466; zirbel@bgsu.edu mailto:zirbel@bgsu.edu http://www-math.bgsu.edu/~zirbel http://www-math.bgsu.edu/%7Ezirbel

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/BGSU-RNA/FR3D/issues/11#issuecomment-278440194, or mute the thread https://github.com/notifications/unsubscribe-auth/AAj8xM56RwR5MYrTnDTzNIskV7RZbR_cks5rahuLgaJpZM4L7Mo2.

BGSU-RNA / FR3D