Technical Support Forum Index
Technical Support Forum
Access ChemAxon scientists and developers here. For confidential or other support please email.
The time now is Mon Dec 01, 2008 6:34 pm
 <b>Register</b> (required to post and download)Register (required to post and download)
 Username:    Password:   Remember login       
 FAQFAQ   SearchSearch   DigestsDigests 
Different SMART Definitions?
To watch this topic for replies  Register (enables digests) or give email address:
Reply to topic    Home -> Forums -> Support -> Library profiling: Virtual screening, clustering & molecular descriptors -> Different SMART Definitions?

Display posts from previous:        View previous topic :: View next topic    
Author Message
fp
Joined: 09 Sep 2004
Posts: 8


View user's profile



Back to top
Post subject: Different SMART Definitions?
Link to postPosted: Wed Sep 15, 2004 6:31 pm  Reply with quote

I think that there are different SMART definitions used for substructure searching in the jsp-example and for searching with the isMatching("") method in the MDGEnerator class.

I have a small database and a "substructure search" with the .jsp example returns two molecules containing "CC(N)(O)c1ccccc1" (search string displayed in the tomcat command prompt).

On the other hand, I set a bit in a custom fingerprint if a call of

private boolean genKey170() { return isMatching( "CC(N)(O)c1ccccc1" ); }

returns true. So this Bit should be set to "1" for two molecules in the database, but it is to 0 for every molecule. Can you tell me why? Until now everything seemed to work fine with the custom fingerprint, the other bits for the molecules seem to be set correctly (perhaps there the interpretation of the SMART-Strings is the same?).
Szabolcs
Joined: 03 Jun 2004
Posts: 822
ChemAxon personnel

View user's profile

Visit poster's website

Back to top
Post subject: Re: Different SMART Definitions?
Link to postPosted: Thu Sep 16, 2004 10:08 am  Reply with quote

Most probably the problem is that your query "CC(N)(O)c1ccccc1" is both valid in smiles and in smarts. The main difference here is the handling of uppercase symbols: when interpreted as smarts, these get an extra aliphatic query flag, and will only match aliphatic atoms. (See figure attached.)

I presume that you pasted this string into the canvas of the marvin applet in the jsp example. This interprets it as smiles because no smarts features are present. If you want to import your string as smarts, you have to follow these steps:
1. select Edit/Source
2. select Format/SMARTS in the pop-up window
3. paste your smarts string into the text area of the window
4. select File/Import SMARTS.

This should give similar results as the code.



 Filename: smarts_smiles.PNG    Filesize: 2.7 KB    Viewed: 10048 Time(s)
 Description:  Difference between smarts and smiles imports.
smarts_smiles.PNG
fp
Joined: 09 Sep 2004
Posts: 8


View user's profile



Back to top
Post subject:
Link to postPosted: Thu Sep 16, 2004 1:29 pm  Reply with quote

Tanks!
Is the isMatching(string1) - method searching for SMARTS String string1 and the jsp-applet is using SMILES then?
mvargyas
Joined: 21 May 2004
Posts: 639
ChemAxon personnel

View user's profile

Visit poster's website

Back to top
Post subject:
Link to postPosted: Thu Sep 16, 2004 1:40 pm  Reply with quote

isMatching() uses SMARTS in the sample code, but actually it is not isMatching() that sets SMARTS mode. Instead, the query structure is read in as SMARTS. You find the corresponding code in the constructor (in the original example):

Code:
public MaccsGenerator() {
        search = new MolSearch();
        smartsReader = new MolHandler();
        smartsReader.setQueryMode( true ); // this makes the trick!!!
        elemAnal = new ElementalAnalyser();
        arom = new Aromata();
    }


The query structure in the JSP example is interpreted as SMILES, because it does not contain any specific query property that could only bbe represented in SMARTS.

Miklos
fp
Joined: 09 Sep 2004
Posts: 8


View user's profile



Back to top
Post subject:
Link to postPosted: Thu Sep 16, 2004 1:59 pm  Reply with quote

thanks, now it's clear. Miklos, have you made any progress with the metric-problem?
mvargyas
Joined: 21 May 2004
Posts: 639
ChemAxon personnel

View user's profile

Visit poster's website

Back to top
Post subject:
Link to postPosted: Thu Sep 16, 2004 2:00 pm  Reply with quote

I sent you an e-mail regarding the metric problem.
Miklos
Display posts from previous:   
Reply to topic    Home -> Forums -> Library profiling: Virtual screening, clustering & molecular descriptors -> Different SMART Definitions? All times are GMT + 1 Hour
Page 1 of 1
To watch this topic for replies   Register (enables digests) or give email address  
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum