bug in knime node 1.1.0.201201291733

[Hinnerk Rey]

Hi,
after installation of latest nodes (Indigo KNIME integration
1.1.0.201201291733) I could not load my sample any more. It always
terminates with a Visual C++ runtime error and takes knime with it
(either molecule to indigo and query molecule to indigo).
Best, Hinnerk

PS: Many thx for the C example of the transformation.

[Hinnerk Rey]

Sorry, my fault. It was again the "write tables to disc" issue. I have
forgotten this.

[Hinnerk Rey]

nevertheless. murcko and scaffold finder are not stable and results in
either crash (Vusial C++ runtime error) or lots of errors in log (both
nodes with write tables to disc, 32bit, win7). I try to figure out at
least one problematic structure.

[Mikhail Rybalkin]

Hello Hinnerk,

Thank you for informing us that the nodes has some instability
problem! We should fix them before releasing the stable version. It
would be great if you could provide us some example. Also could you
tell how much RAM do you have on your computer, and how many molecules
are being proceeded?

Mikhail

[Hinnerk Rey]

Hi,
I have 8gb ram and this sample is ~100k compounds. But i get the
errors quite early. Here is the log:
**************************************************************************
ERROR Murcko Scaffold Execute failed: pool: access to unused element
1
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
ERROR BufferFromFileIteratorVersion20 Errors while reading row 2152
from file "knime_container_20120131_919338784171118525.bin.gz": cis-
trans: can't sort restored substituents; Suppressing further warnings.
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring
ERROR Scaffold Finder Execute failed: array: reserve(): no memory
ERROR SyncExecQueueDispatcher Uncaught exception while queuing
events into main thread

Row 2152 is empty.

Best, Hinnerk

On Jan 30, 10:47 pm, Mikhail Rybalkin <rybal...@ggasoftware.com>
wrote:

[Mikhail Rybalkin]

Hello Hinnerk,

These error messages are really not expected, and something has gone
wrong. Could you update the Indigo nodes: I have added an addional
diagnostics, by printing rowid for the row that caused this error. In
particular, I would like to know what molecule caused "cis-trans:
can't sort restored substituents" and "pool: access to unused element"
messages.

And are you using 32-bit or 64-bit KNIME?

Best,
Mikhail

[Hinnerk Rey]

Hi Mikhail,
I use 32bit with the latest updates / nodes (Indigo KNIME integration
1.1.0.201201312226). Knime is freshly started without any other nodes
(only SDFile Reader and Indigo-Nodes).
I still get the same error (row 2152 has a missing cell in the MOL-
Block). My 64bit installation with an old version (how do i determine,
what is installed) works OK.

ERROR Murcko Scaffold Execute failed: pool: access to unused element
1
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
ERROR BufferFromFileIteratorVersion20 Errors while reading row 2152

from file "knime_container_20120201_4069121090868567364.bin.gz": cis-

trans: can't sort restored substituents; Suppressing further warnings.
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring
ERROR IndigoMolCell Error while unserializing Indigo object: cis-
trans: can't sort restored substituents
WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring

ERROR IndigoMolCell Error while unserializing Indigo object: array:
reserve(): no memory

WARN IndigoScaffoldFinderNodeModel Molecule table contains missing
cells: ignoring

[Mikhail Rybalkin]

Hello Hinnerk,

As you have the errors only in the 32-bit version, I think that the
problem is with the memory management inside our nodes in Knime. We
will try to fix it after returning from the 5th KNIME Users Group
Meeting and Workshop.

Just to ensure that the 64-bit version works fine, could you check the
same workflow with the 64-bit Knime with the latest indigo nodes?

Best regards,
Mikhail

[Hinnerk Rey]

Hi Mikhail,
with 64bit and the latest nodes i also get:

ERROR Murcko Scaffold Execute failed: pool: access to unused element

2
WARN IndigoScaffoldFinderNodeModel exact method has reached
iteration limit: trying to search approximate
ERROR Scaffold Finder Execute failed: element: bad valence on B
having 5 drawn bonds, charge -2, and 0 radical electrons

Best, Hinnerk
PS: Have fun on the UGM

[Savelyev Alexander]

Dear Hinnerk,

We have uploaded a new KNIME nodes version. The new version is much
more stable to huge data sets. Also, some runtime crashes were fixed.
Could you download the new version from the nightly builds. It would
be very helpful for us to know about the new behaviour with huge data
inputs. Please let us know if any exceptions are raised.

With best regards,
Alexander

[Hinnerk Rey]

Hi
I can isolated the problem. There are 2. First, the node "molecule to
indigo" does not work correct with moil-files any more. With "Indigo
KNIME integration 1.1.0.201111071836" (knime 64bit, 2.5.0, win7) I can
import rows with a molfile. Now with the latest version, I always get
"missing cell", although the renderer can render the rows (I see a
picture in SDFreader). Furthermore with the mentioned version I can
process attached entry. This means, the entry shows up in invalid
molecules, which is not the best behavior, but OK. Now the entry leads
to a crash of the node, that means no processing of any row is
possible. Furthermore any processing of a missing cell with an indigo
node can have strange side effects.
I hope, this helps to identify some of the problems.
Best, Hinnerk

------------------------------------------------

test charge -14

52 72 0 0 0 0 999 V2000
3.9020 1.9590 2.0320 W 0 0 0 0 0 7 0 0 0 0 0 0
2.8490 1.9830 1.9530 O 0 0 0 0 0 3 0 0 0 0 0 0
3.9280 2.9290 2.0940 O 0 0 0 0 0 0 0 0 0 0 0 0
3.9110 1.8860 3.0400 O 0 0 0 0 0 0 0 0 0 0 0 0
3.8940 2.0430 1.0280 O 0 0 0 0 0 0 0 0 0 0 0 0
3.8700 0.9360 1.9420 O 0 0 0 0 0 0 0 0 0 0 0 0
4.9170 1.9230 2.0340 O 0 0 0 0 0 0 0 0 0 0 0 0
2.9070 2.0620 1.0020 W 0 0 0 0 0 7 0 0 0 0 0 0
2.8450 0.9870 1.9270 W 0 0 0 0 0 7 0 0 0 0 0 0
3.4710 3.6900 2.5270 W 0 0 0 0 0 7 0 0 0 0 0 0
3.3700 2.3040 3.8450 W 0 0 0 0 0 7 0 0 0 0 0 0
2.8730 1.0860 0.9360 O 0 0 0 0 0 0 0 0 0 0 0 0
1.9210 2.0900 0.9770 O 0 0 0 0 0 0 0 0 0 0 0 0
2.9390 3.0440 1.0750 O 0 0 0 0 0 0 0 0 0 0 0 0
2.9240 2.1410 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
2.8210 0.8970 2.9140 O 0 0 0 0 0 0 0 0 0 0 0 0
1.8260 1.0470 1.9120 O 0 0 0 0 0 0 0 0 0 0 0 0
2.8020 0.0000 1.8370 O 0 0 0 0 0 0 0 0 0 0 0 0
2.5790 3.3100 2.3340 O 0 0 0 0 0 3 0 0 0 0 0 0
3.6260 3.2290 3.4460 O 0 0 0 0 0 0 0 0 0 0 0 0
3.4390 4.1820 1.6500 O 0 0 0 0 0 0 0 0 0 0 0 0
3.0170 4.4430 2.9630 O 0 0 0 0 0 0 0 0 0 0 0 0
4.3550 4.0930 2.7120 O 0 0 0 0 0 0 0 0 0 0 0 0
2.4950 2.2680 3.3130 O 0 0 0 0 0 3 0 0 0 0 0 0
3.1170 1.3800 4.2350 O 0 0 0 0 0 0 0 0 0 0 0 0
2.8710 2.7790 4.6530 O 0 0 0 0 0 0 0 0 0 0 0 0
4.2500 2.3110 4.3960 O 0 0 0 0 0 0 0 0 0 0 0 0
1.1700 2.5530 1.3630 W 0 0 0 0 0 7 0 0 0 0 0 0
2.5400 3.7780 1.4810 W 0 0 0 0 0 7 0 0 0 0 0 0
2.2380 1.3690 3.7220 W 0 0 0 0 0 7 0 0 0 0 0 0
1.0320 1.5320 2.3060 W 0 0 0 0 0 7 0 0 0 0 0 0
2.1500 4.0740 2.7720 W 0 0 0 0 0 7 0 0 0 0 0 0
1.9760 2.7490 4.0910 W 0 0 0 0 0 7 0 0 0 0 0 0
1.5220 2.3840 2.2620 O 0 0 0 0 0 3 0 0 0 0 0 0
1.6480 3.3810 1.3130 O 0 0 0 0 0 0 0 0 0 0 0 0
0.6960 1.7150 1.4030 O 0 0 0 0 0 0 0 0 0 0 0 0
0.4230 3.0080 1.7400 O 0 0 0 0 0 0 0 0 0 0 0 0
0.8220 2.7140 0.4580 O 0 0 0 0 0 0 0 0 0 0 0 0
2.1260 4.5370 1.9380 O 0 0 0 0 0 0 0 0 0 0 0 0
2.5000 4.2560 0.6350 O 0 0 0 0 0 0 0 0 0 0 0 0
1.3710 1.3410 3.2080 O 0 0 0 0 0 0 0 0 0 0 0 0
1.7720 1.8220 4.4860 O 0 0 0 0 0 0 0 0 0 0 0 0
2.0080 0.4380 4.1300 O 0 0 0 0 0 0 0 0 0 0 0 0
0.2410 2.0150 2.6930 O 0 0 0 0 0 0 0 0 0 0 0 0
0.5200 0.6890 2.3390 O 0 0 0 0 0 0 0 0 0 0 0 0
2.1930 3.6640 3.7030 O 0 0 0 0 0 0 0 0 0 0 0 0
1.2810 3.7070 2.5900 O 0 0 0 0 0 0 0 0 0 0 0 0
1.7150 4.8420 3.2140 O 0 0 0 0 0 0 0 0 0 0 0 0
1.0990 2.7180 3.5340 O 0 0 0 0 0 0 0 0 0 0 0 0
1.4930 3.1990 4.8720 O 0 0 0 0 0 0 0 0 0 0 0 0
0.7620 2.8670 2.6390 W 0 0 0 0 0 7 0 0 0 0 0 0
0.0000 3.3400 3.0140 O 0 0 0 0 0 0 0 0 0 0 0 0
1 2 1 0 0 0 0
1 3 1 0 0 0 0
1 4 1 0 0 0 0
1 5 1 0 0 0 0
1 6 1 0 0 0 0
1 7 2 0 0 0 0
2 8 1 0 0 0 0
2 9 1 0 0 0 0
3 10 1 0 0 0 0
4 11 1 0 0 0 0
5 8 1 0 0 0 0
6 9 1 0 0 0 0
8 12 1 0 0 0 0
8 13 1 0 0 0 0
8 14 1 0 0 0 0
8 15 2 0 0 0 0
9 12 1 0 0 0 0
9 16 1 0 0 0 0
9 17 1 0 0 0 0
9 18 2 0 0 0 0
10 19 1 0 0 0 0
10 20 1 0 0 0 0
10 21 1 0 0 0 0
10 22 1 0 0 0 0
10 23 2 0 0 0 0
11 20 1 0 0 0 0
11 24 1 0 0 0 0
11 25 1 0 0 0 0
11 26 1 0 0 0 0
11 27 2 0 0 0 0
13 28 1 0 0 0 0
14 29 1 0 0 0 0
16 30 1 0 0 0 0
17 31 1 0 0 0 0
19 29 1 0 0 0 0
19 32 1 0 0 0 0
21 29 1 0 0 0 0
22 32 1 0 0 0 0
24 30 1 0 0 0 0
24 33 1 0 0 0 0
25 30 1 0 0 0 0
26 33 1 0 0 0 0
28 34 1 0 0 0 0
28 35 1 0 0 0 0
28 36 1 0 0 0 0
28 37 1 0 0 0 0
28 38 2 0 0 0 0
29 35 1 0 0 0 0
29 39 1 0 0 0 0
29 40 2 0 0 0 0
30 41 1 0 0 0 0
30 42 1 0 0 0 0
30 43 2 0 0 0 0
31 34 1 0 0 0 0
31 36 1 0 0 0 0
31 41 1 0 0 0 0
31 44 1 0 0 0 0
31 45 2 0 0 0 0
32 39 1 0 0 0 0
32 46 1 0 0 0 0
32 47 1 0 0 0 0
32 48 2 0 0 0 0
33 42 1 0 0 0 0
33 46 1 0 0 0 0
33 49 1 0 0 0 0
33 50 2 0 0 0 0
34 51 1 0 0 0 0
37 51 1 0 0 0 0
44 51 1 0 0 0 0
47 51 1 0 0 0 0
49 51 1 0 0 0 0
51 52 2 0 0 0 0
M CHG 1 51 -14
M END
$$$$
------------------------------------------------




On Feb 18, 7:02 pm, Savelyev Alexander <asavel...@ggasoftware.com>
wrote:

[Savelyev Alexander]

Dear Hinnerk,

Thank you for your attention. I have reproduced the bug. It seems like
changes have not been so painless, as we would like to. Now the
serialization errors reveal. I think the issue was not noticed in the
old version, because cells were saved only if memory ended (or if the
"write tables to disc" flag is specified). Anyway, I have fixed the
crashes. Now, a verification for input structures is applied. We have
uploaded the new version, and it will be available soon in nightly
builds.
Could you also specify the nodes, where missing cells lead to the
"strange side effects".

Best regards,
Alexander

[Hinnerk Rey]

Hi Alexander,
reading mol files with Molecule to Indigo is now OK. When I try to
detect Murcko scaffolds, I immediately get "ERROR Murcko Scaffold
Execute failed: pool: access to unused element 1".
When I convert the mol files to smiles with ChemAxon MolConvert-Node,
import the smiles to Indigo, and detect Murcko, it works.
If i try to convert mol to smiles wit Indigo via import / export i get
a runtime exception.
Trying the MCS-node with smiles as sources leads to "ERROR
IndigoScaffoldFinderNodeModel internal error while launching extract
scaffold: array: reserve(): no memory
"
Best, Hinnerk



On Feb 21, 4:08 pm, Savelyev Alexander <asavel...@ggasoftware.com>
wrote:

[Savelyev Alexander]

Dear Hinnerk,
Many thanks for the given examples. We have fixed the molecule loading
errors. New features were added to the KNIME nodes and to the internal
format for avoiding such crashes. The new version is available in the
nightly builds (1.1.0.201202272004)
As for the last issues
- Murcko scaffold: could you give an example molecules?
- MCS: how many molecules are processed?. Could you also specify the
memory parameters (JVM settings and nodes memory policy)

With best regards,
Alexander

[Hinnerk Rey]

Hi Alexander,
many thx for the quick fix of the valence problem. This seems to work
now (I do not get any analog errors any more).
Here is a mol, which causes the problem with murcko in the worklfow,
sdfreader, indigo2molecule, indigo-murcko (ERROR Murcko Scaffold
Execute failed: pool: access to unused element 1).
With MCS. Maybe my set was to large. :-) I will check this later.
Best, Hinnerk

--------------------

test

19 19 0 0 0 0 0 0 0 0999 V2000
4.0950 2.2596 0.0000 C 0 0 2 0 0 0 0 0 0 0 0 0
3.2858 2.8467 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
3.7859 1.3083 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.5951 3.1258 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.9242 1.7004 0.0000 C 0 0 2 0 0 0 0 0 0 0 0 0
2.4766 2.2596 0.0000 C 0 0 1 0 0 0 0 0 0 0 0 0
2.7857 1.3083 0.0000 C 0 0 1 0 0 0 0 0 0 0 0 0
4.0950 3.9920 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
5.5954 3.1258 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
4.5981 0.7552 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
5.8384 1.2943 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
1.9765 3.1258 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
1.5644 1.8525 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
2.1975 0.4991 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.5951 4.8582 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
2.4766 3.9920 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
0.9772 3.1258 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
1.9775 4.8582 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
5.4093 2.5746 0.0000 H 0 0 0 0 0 0 0 0 0 0 0 0

1 2 1 0 0 0 0
1 3 1 0 0 0 0

1 4 1 6 0 0 0

1 5 1 0 0 0 0

2 6 1 0 0 0 0
3 7 1 0 0 0 0
4 8 1 0 0 0 0
4 9 2 0 0 0 0
5 10 1 0 0 0 0
5 11 1 0 0 0 0
5 19 1 1 0 0 0
6 7 1 0 0 0 0
6 12 1 0 0 0 0
6 13 1 1 0 0 0
7 14 1 1 0 0 0
8 15 1 0 0 0 0
12 16 1 0 0 0 0

12 17 2 0 0 0 0

16 18 1 0 0 0 0
M STY 3 1 DAT 2 DAT 3 DAT
M SAL 1 1 1
M SDT 1 MDLBG_STEREO_KEY F
M SDD 1 3.5950 1.7996 DA ALL 1 5
M SED 1 R
M SAL 2 1 5
M SDT 2 MDLBG_STEREO_KEY F
M SDD 2 4.4242 1.2404 DA ALL 1 5
M SED 2 S
M SAL 3 1 6
M SDT 3 MDLBG_STEREO_KEY F
M SDD 3 1.9766 1.7996 DA ALL 1 5
M SED 3 R
M END

$$$$
--------------------------

On Feb 28, 11:25 am, Savelyev Alexander <asavel...@ggasoftware.com>
wrote:

[Hinnerk Rey]

Hi,
here the next problem. :-)
When i open a file (ReadFile node) with 10k entries
(B1=CB=c2cc3B=CC=c3cc12 each line), i can cast them to smiles and can
import them to Indigo. While exporting these structures to InChIKey,
Knime vanishes (crash wihout any error, etc). I do not know, whether
this is a Indigo problem and / or Knime problem.
Best, Hinnerk

Environment: Win7 64bit, 8gb ram, Knime 2.5.2 32bit, latest Indigo
nodes.

[Mikhail Rybalkin]

Hello  Hinnerk,

We have fixed this issue in the latest nightly build. Thank you for the bug report!

Best regards,

Mikhail

[Hinnerk Rey]

Hi Mikhail,
I can confirm, that with latest nodes, the smile-inchi problem does
not further crash knime. Nevertheless, the result is not as expected
(Could not convert structure with RowId='Row1': core: Indigo-InChI:
InChI generation failed: (null). Code: 2.). If this problem is somehow
caused by InChI-sources, let me know. Next week we have a meeting.
And do you have a solution for the reported murcko-problem?
Best, Hinnerk

[Mikhail Rybalkin]

Hi Hinnerk,

I have fixed an issue with murcko problem. You can check this in the latest Indigo Knime nodes 1.1.3.

And also I have restored the error message notification as in the previous version  there was a "(null)" instead of an error message. For the molecule provided it gives "Accepted unusual valence(s): C(3); Cannot process aromatic bonds. Code: 2." I will check why it is so.

Just to make clear I want to say that we do not compute a stereocenter layout in the InChI yet, but this is planned to be done soon. How is it important for your application?

Best regards,
Mikhail

[Hinnerk Rey]

Hi Mikhail,
InChI is for me not that important, since in an productive environment
I rely on the InChI binaries (comparison with other databases). I
combination with indigo (via c or knime) it is an easy way for me to
use InChI in an early stage of development. I know your limitation
with stereo and the general InChI limits (inorganic /
organometallics).
Best, Hinnerk

PS: Is the InChI dll / so OK for you, or would you prefer a java
interfaces directly provided by InChI?

[Hinnerk Rey]

Hi Mikhail,
just for info. Murcko is now OK.
Best, Hinnerk

[Mikhail Rybalkin]

Hi Hinnerk,

> PS: Is the InChI dll / so OK for you, or would you prefer a java interfaces directly provided by InChI? 

We are using InChI source code to create a indigo-inchi.dll/so/dylib for all platforms, and this Indigo InChI plugin used not only in Knime via Java interface, but also it can be used in Python or C#. So we are not using dynamic libraries provided by InChI because we need to support different platforms and programming interfaces via single Indigo InChI plugin for Indigo. Am I right, that you are asking as a developer how to make InChI programming interface better? Because only now I have found that you are a member of InChI subcommittee.

Best regards,

Mikhail

[Mikhail Rybalkin]

Also I have checked the SMILES B1=CB=c2cc3B=CC=c3cc12 converted into molfile with the winchi-1.exe InChI utility and it works fine. I think that there might be some problem in providing correction information about valencies from Indigo into InChI library, and I will check this.

[Hinnerk Rey]

> Am I right, that you are asking as a developer how to make InChI
> programming interface better?

That was my question. I know, that you can access inchi via c and
languages, which support a c interface (python, .net, etc.). But in
the past, there have been the demand to support a java interface,
since various chemical information tools only support java. I know,
that there is a (slow?) inchi-jni solution, a chemaxon solution, etc.
They all have the disadvantage of not being officially supported.
Therefore the question, if you would also prefer a java interface or
if you are fine with the existing libraries.


On Mar 14, 11:00 am, Mikhail Rybalkin <rybal...@ggasoftware.com>

> ...
>
> read more »

[Mikhail Rybalkin]

We are providing common interface for Indigo functionality in different languages, and so we do not need InChI Java interface. Even more it will bring to us more problems, because we cannot use InChI Java library for .NET or Python. So the short answer is that for our purpose we can use only the InChI source code. But if someone developing an application in Java then the official InChI Java library will be very helpful.

[Hinnerk Rey]

Hi Mikhail,
coming back to the topic.
Following mol file cannot be displayed a the valence check ends with
an error (ERROR Valence Checker Execute failed: cis-trans: can't
sort restored substituents).
Best, Hinnerk


--------------------

test

13 14 0 0 0 0 0 0 0 0999 V2000
3.2280 1.9670 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
3.2280 2.9650 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.1770 1.6580 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
2.3640 1.4650 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.1770 3.2730 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
2.3600 3.4680 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.7640 2.4660 0.0000 N 0 0 0 0 0 0 0 0 0 0 0 0
4.4860 0.7070 0.0000 N 0 0 0 0 0 0 0 0 0 0 0 0
1.4930 1.9680 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
4.4860 4.2240 0.0000 N 0 0 0 0 0 0 0 0 0 0 0 0
1.4930 2.9700 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0
5.4640 0.5000 0.0000 Br 0 0 0 0 0 0 0 0 0 0 0 0
5.4640 4.4320 0.0000 H 0 0 0 0 0 0 0 0 0 0 0 0

1 2 1 0 0 0 0
1 3 1 0 0 0 0

1 4 2 0 0 0 0
2 5 1 0 0 0 0
2 6 2 0 0 0 0
3 7 2 0 0 0 0
3 8 1 0 0 0 0
4 9 1 0 0 0 0
5 7 1 0 0 0 0
5 10 2 3 0 0 0
6 11 1 0 0 0 0
8 12 1 0 0 0 0
9 11 2 0 0 0 0
10 13 1 0 0 0 0
M END
-----------------------------b