Chemaxon Bcut Descriptors for NCI & L2 Discrepancy Analy

Recently I developed a machine learning algorithm called L2 Discrepancy Learning .





I applied the work to Mine High-Throughput Screening for Drug Discovery. I used the data AIDS antiviral activity Data of National Cancer Institute. These compounds are involved in the treatment of HIV infection by: Reverse transcriptase (RT) inhibitors, Protease inhibitors (PI) and Fusion inhibitors .





I used Bcut descriptors for compounds( year 1999) NOT generated by Chemaxon.








The main results are in:


http://faysal.el.khettabi.googlepages.com/applicationtodtp-nciaidsantiviralscreen





For more analysis of the L2 Discrepancy framework, I would like to


have the Chemaxon Bcut descriptors for compounds


(year 1999 or year 2004).





If someone has this data, please let me know.





thank you very much.





Faysal.

I don't quite understand your question:


do you ask for the compounds (year 1999 or year 2004) or the related BCUT values.


We don't have the compounds, but if you can access them, the our generatemd tools can assign various BCUT descriptors to the structures.





Does this help at all?





Regards,


Miklos

Thanks for the replay.





What I mean:


Compounds of HIV and Cancer Compounds have been generated under Anti-HIV Screening Data and Anti-Cancer Screening Data program, see the link:





http://dtp.nci.nih.gov/screening.html





AIDS( year 1999) Compounds are grouped in:


http://dtp.nci.nih.gov/docs/aids/aids_data.html





Cancer( year 2006) Compounds are grouped in:


http://dtp.nci.nih.gov/docs/cancer/cancer_data.html





A set of unified descriptors(bcut for instance) for all the compounds in "DTP - AIDS Antiviral Screen" will be a great deal to stem a close collaboration of data mining scientists, biologists, and computational chemists. Also it will provide a venue to facilitate the exchange of ideas between the three disciplines.





It will be suitable that only a leading professional organization of computational chemistry generate the descriptors. The data will be considered as an online reference data that will provide a premier forum for advancement and adoption of the "science of knowledge discovery and data mining" to the drug discovery.


With my best regards,


Faysal.