Are all these structures in the considered differently?

User 52a4e280f0

17-09-2012 11:55:36

Hi,


I am trying to load few articles which have got some structures in the form or sdf or mol files.


I also have duplicate filtering enabled in my relational database.


When I load the article set for the first time, few structures get created in the structure table.


When I load the same set again, one new structure gets added to the table. Although I was not expecting this to happen.


I have attached the file which I exported from my databse. This contain all the structures which got re-inserted every time I reloaded the same set of articles.


Can you please have a look


Many thanks,


Isha

ChemAxon abe887c64e

17-09-2012 15:28:59

Hi Isha,


Could you tell us which version of JChem are you using?


Thanks,


Krisztina

User 52a4e280f0

18-09-2012 08:28:29

Hi Kristina,


I am using version 5.10.4


Thanks,


Isha

User 52a4e280f0

18-09-2012 09:05:40

I have attached below the sdf file which every time upon getting loaded creates a new structure in the relational database.


Not sure why??

ChemAxon abe887c64e

18-09-2012 09:14:58

Thank you the data. This operation is really erroneous. We start to find the reason, and will correct as soon as possible.


Best regards,


Krisztina

User 52a4e280f0

25-09-2012 08:52:27

Hi Kriztina,


Can you please let me know if there is any update on this?


Thanks,


Isha

ChemAxon abe887c64e

25-09-2012 14:15:01

Hi Isha,


Unfortunately, the problem seems to be more complex than it seem'd to be at the first reproduction. We could not solve it yet, it needs to be prioritized and scheduled into our workflow.


We would appreciate your help to give us further information:


How urgent would be the solution for you? Did you find this bug in production or in evaluation environment?


Do you have any other workaround to filter out duplicate structures? If not, you can try to import in "Markush library" type table; we found that duplicate filtering of the attached EXPORTEDDATA.mrv works better in Markush table.


Best regards,


Krisztina


 

User 52a4e280f0

27-09-2012 15:06:44

Thanks Krisztina for the update. The concerned team at our end will be looking into this to analyse the impact


of this on our seach results and will let us know if and when this needs to be prioritised. 


I have also recieved a separate explanation in www.chemaxon.com%2fforum%2fviewtopic.php%3fp%3d44582%2344582" target="_blank" style="font-family: monospace; font-size: 13px;">https://www.chemaxon.com/forum/viewtopic.php?p=44582#44582 and have forwarded this to the concerned people.


Isha

User 52a4e280f0

27-09-2012 15:09:19

Also you suggested to use "Markush library" type table.


Presently we use ANY type table. 


Can you please explain how different are they from each other and if we replace ANY with Markush what sort of changes would we be looking at?

ChemAxon abe887c64e

28-09-2012 08:54:13

Hi Isha,


Please find information about Markush library type tables in the following documents:


http://www.chemaxon.com/jchem/doc/dev/dbconcepts/index.html#tabletypes


http://www.chemaxon.com/jchem/doc/user/query_markush.html


Best regards,


Krisztina

User 52a4e280f0

28-09-2012 10:56:29

Hi Krisztina,


It looks like we will need to stick with the ANY table type and cannot go with the MARKUSH ones as we get lot of different types of structures.


Can you please give us an estimated timeline for when do you think this bug could be fixed at the earliest as it seems to be affecting our search results?


When we are running an EXACT match for such structures, it only returns the last one created and so we only get to see the content which has been mapped to it, whereas we expected all the structures to be returned while running an exact match search?


So, it appears the bug is while loading the structure? While searching it only returns the last cd_id created?


Thanks,


Isha 

ChemAxon abe887c64e

01-10-2012 11:08:29

Hi Isha,


We start to fix the found bugs, and hopefully build the fix in the next patch release. However, could you send us other example structures of erroneous duplicate filtering ? We would like to see preferably all the structure features that may cause the problem.


Thanks,


Krisztina

ChemAxon fb166edcbd

08-10-2012 13:16:19

We have fixed this problem, the correction will be available in the next minor (5.11.2) and major (5.12) releases.


The problem was related to initialization of our search mechanism.


Thanks for the report.


Could you check if the other problem you have reported:


https://www.chemaxon.com/forum/ftopic9552.html


is the same as this one?


Can we close this issue?

User 52a4e280f0

08-10-2012 13:45:30

Thanks Nora...


 


I think both of them are related. I will wait for fixed release to see if it works at my end.


 


When is it supposed to be released?


 


Thanks,


Isha

ChemAxon fb166edcbd

08-10-2012 14:30:34

The 5.11.2 release will be out in 2 weeks time.