Jcsearch settings matching the GUI’s default

Hallo,

My GUI Instant JChem  version is 15.4.27.0 on OSX 10.10.3. My command line (jcsearch) versions are 15.3.9.0 on OsX 10.10.3 and 15.4.6.0 on Ubuntu 14.04.

The problem is that by visual inspection the and by using the default settings of the GUI version I am able to find the substructure from this smile "CCCCNCC"  in this structure "CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1".

However, the following command line does not produce any result on neither of the jcsearch versions mentioned above.

> jcsearch -q "CCCCNCC"  "CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1"

So I would like to ask:
What  “–flags" should I set in any of the mentioned command line versions so their output would match the GUI's version (default settings) output?

 
Would it help me if I had a newer version of the command line tools and where can I get it from?

Thank you for answer in advance!

Best

Tsonko

Dear Tsonko,

I am sorry for delayed answer. By default, IJC uses similarity search with the settings that you can see in the attached screenshot. The relevant command line arguments to match this settings can be found here. There should not be any need to download newer JChem suite (as you refer to as command line tools) as you are already on version 15.3.9.0 and later, which have the same JChem table versions and are compatible. 

 

I hope this helps and wishing all the best

David

Dear David,

 

Thank you for your time and answer!

However, unfortunately your answer does not help.

I am well aware where the settings are and I have been fighting
with them for the last 3 weeks.

 

I came to the conclusion that we might talk about a big
there.

Please try the following code:

> jcsearch -q "CCCCNC"
"CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1"

You suppose to get an output :

CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1

 

If you however add an extra carbon to the query

jcsearch -q "CCCCNCC" "CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1"

You would not get any output.

 

Now the funny part is the following:

If you put the second query to a text file (I have provided
one) and use it in the same fashion you should get the same output as from the
first query.

> jcsearch -q 
query.smi "CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1"

Output:

CCCCN1C(Nc2ccccc2C1=O)c1ccc(OC)c(COc2ccc(NC(C)=O)cc2)c1

 

So to sum up:

The jcsearch works with some smiles (CCCCNC) as it should
and does not work with others (CCCCNCC).

The smile CCCCNCC is part of the molecule and supposes to produce
the same result as the first one when used in the same substructure search as a
query

The smile CCCCNCC  produces
result if it is recorded in a text file.

 

My questions are:

Does the jcsearch suppose to behave in such a way?

And if yes,

How to construct a command line with smiles (like
the first example) in a way that will be consistent so the program would
produce correct results.

 

I Thank you for your time and I am looking forward for your answer!

 

Best

Tsonko

Hello Tsonko,

I've reproduced your issue and reported it to JChem base team, I will tell you once this issue gets resolved.

thanks for your time, and this well written bug-report.

Best regards,

Martin

Hello Tsonko,

I received an answer from Jchem Base team, the reason why the jcsearch is was not finding the SMILES is because it was expecting the query string in SMARTS (difference is that C in SMILES means carbon atom (aromatic or aliphatic), while in SMARTS this is exclusively aliphatic carbon) 

to use SMILES in jcsearch, you can add --tautomer so the search string will be used as SMILES (instead of SMARTS) you can also change the query to CCCCNCc which would match the molecule from your example. I dont think Instant JChem has sufficiently straightforward way on how to search structures by SMARTS. easiest way how to search for the same structure described by the SMARTS as jcsearch does is to add explicit hydrogen atoms as shown in the attached picture.

Best regards,

Martin

Hallo Martin,

Thank you for clarifying that issue for me. It have not cross my mind that my query was not a SMILES but SMARTS :).

Lots of things to be learned I guess :).

Thank you again for the help!

Best

Tsonko