optimize screening

User 1033cfd7dc

24-01-2007 13:14:37

Hello,





when I run the example (NCI1000, ACE) of optimize.bat, I got a results with enrichment=1, selectivity=0.5 for all cases. I don't know how to correct it. I even don't know whether the hitstatisctics or optimizemetrics is the origin of this problem. The optimizemetrics itself returns resonable values in the dos prompt.





See the output below. Could you tell me what to do?





Zsolt











.


.


.


. Usage: optimize {target file} {actives file}


. Target and actives can be chosen from the directory


. 'examples\molecules', e.g.:


. optimize nci500.smiles beta2_adrenoceptor_antagonists.smiles


.


.


. This example generates one set of molecular files for parameter


. optimization and evaluation of different parametrized metrics of two


. molecular descriptors (pharmacophore fingerprints and chemical


. fingerprints). Then it performs the optimization of parameters and


. evaluation of the performance of the generated parametrized metrics.


. These metrics are used in screening a set of molecules against a set of


. known actives.


.


.


. Results are written into the file example_statistics.stat.


.


.





RandomMS - Select subset of molecules, (C) 2002-2005 ChemAxon Ltd.





Reading input file nci1000.smiles


Initializing output file opt-target


Initializing output file hit-target


Selecting molecule nr.16


Selecting molecule nr.40


Selecting molecule nr.57


Selecting molecule nr.72


Selecting molecule nr.98


Selecting molecule nr.118


Selecting molecule nr.121


Selecting molecule nr.154


Selecting molecule nr.166


Selecting molecule nr.188


Selecting molecule nr.217


Selecting molecule nr.234


Selecting molecule nr.256


Selecting molecule nr.263


Selecting molecule nr.288


Selecting molecule nr.302


Selecting molecule nr.335


Selecting molecule nr.351


Selecting molecule nr.377


Selecting molecule nr.400


Selecting molecule nr.408


Selecting molecule nr.428


Selecting molecule nr.446


Selecting molecule nr.480


Selecting molecule nr.489


Selecting molecule nr.519


Selecting molecule nr.523


Selecting molecule nr.553


Selecting molecule nr.571


Selecting molecule nr.586


Selecting molecule nr.610


Selecting molecule nr.640


Selecting molecule nr.657


Selecting molecule nr.669


Selecting molecule nr.685


Selecting molecule nr.706


Selecting molecule nr.724


Selecting molecule nr.750


Selecting molecule nr.775


Selecting molecule nr.790


Selecting molecule nr.811


Selecting molecule nr.828


Selecting molecule nr.852


Selecting molecule nr.879


Selecting molecule nr.896


Selecting molecule nr.908


Selecting molecule nr.927


Selecting molecule nr.948


Selecting molecule nr.978


Selecting molecule nr.981


done





RandomMS - Select subset of molecules, (C) 2002-2005 ChemAxon Ltd.





Reading input file ace.smiles


Initializing output file hit-test


Initializing output file opt-actives


Selecting molecule nr.2


Selecting molecule nr.5


Selecting molecule nr.7


Selecting molecule nr.10


Selecting molecule nr.13


Selecting molecule nr.17


Selecting molecule nr.19


done





RandomMS - Select subset of molecules, (C) 2002-2005 ChemAxon Ltd.





Reading input file opt-actives


Initializing output file opt-test


Initializing output file opt-query


Selecting molecule nr.2


Selecting molecule nr.4


Selecting molecule nr.6


Selecting molecule nr.7


Selecting molecule nr.10


Selecting molecule nr.12


done





OptimizeMetrics - Molecular Descriptor Dissimilarity Metrics Optimizer 3.1.6,


(C) 2002-2005 ChemAxon Ltd.





Using Minimum hypothesis





Processing started at Wed Jan 24 13:57:50 CET 2007





Optimizing parameters for descriptor CF, metric Tanimotot


Maximal Enrichment reached 5.185185


Number of similar hits 5


Number of target hits 4


Threshold 0.68786126





Optimizing parameters for descriptor CF, metric Tanimotoa (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.27831602





Optimizing parameters for descriptor CF, metric Euclideant


Maximal Enrichment reached 1.7283951


Number of similar hits 5


Number of target hits 22


Threshold 9.380832





Optimizing parameters for descriptor CF, metric Euclideann (normalized)


Maximal Enrichment reached 4.6666665


Number of similar hits 5


Number of target hits 5


Threshold 0.53209555





Optimizing parameters for descriptor CF, metric Euclideana (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 3.8568122





Optimizing parameters for descriptor PF, metric Tanimotot


Maximal Enrichment reached 1.5053763


Number of similar hits 5


Number of target hits 26


Threshold 0.9221902





Optimizing parameters for descriptor PF, metric Tanimotos (scaled)


Maximal Enrichment reached 1.7948718


Number of similar hits 5


Number of target hits 21


Threshold 0.7437568





Optimizing parameters for descriptor PF, metric Tanimotoa (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Tanimotosa (scaled, asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.28300166





Optimizing parameters for descriptor PF, metric Euclideant


Maximal Enrichment reached 1.0181818


Number of similar hits 6


Number of target hits 49


Threshold 104.26888





Optimizing parameters for descriptor PF, metric Euclideann (normalized)


Maximal Enrichment reached 1.2280701


Number of similar hits 5


Number of target hits 33


Threshold 0.8047408





Optimizing parameters for descriptor PF, metric Euclideanwn (weighted, normalized)


Maximal Enrichment reached 3.1111112


Number of similar hits 5


Number of target hits 10


Threshold 0.03730171





Optimizing parameters for descriptor PF, metric Euclideana (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Euclideanwa (weighted, asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Euclideanan (asymmetric, normalized)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Euclideanwan (weighted, asymmetric, normalized)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Tanimotot0.7


Maximal Enrichment reached 1.5053763


Number of similar hits 5


Number of target hits 26


Threshold 0.9015895





Optimizing parameters for descriptor PF, metric Tanimotos0.7 (scaled)


Maximal Enrichment reached 1.5555556


Number of similar hits 5


Number of target hits 25


Threshold 0.79437464





Optimizing parameters for descriptor PF, metric Tanimotoa0.7 (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.31416464





Optimizing parameters for descriptor PF, metric Tanimotosa0.7 (scaled, asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.14279479





Optimizing parameters for descriptor PF, metric Euclideant0.7


Maximal Enrichment reached 1.0181818


Number of similar hits 6


Number of target hits 49


Threshold 59.224255





Optimizing parameters for descriptor PF, metric Euclideann0.7 (normalized)


Maximal Enrichment reached 1.2962962


Number of similar hits 5


Number of target hits 31


Threshold 0.75422984





Optimizing parameters for descriptor PF, metric Euclideanwn0.7 (weighted, normalized)


Maximal Enrichment reached 4.6666665


Number of similar hits 5


Number of target hits 5


Threshold 0.051405147





Optimizing parameters for descriptor PF, metric Euclideana0.7 (asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 6


Number of target hits 0


Threshold 1.6554015





Optimizing parameters for descriptor PF, metric Euclideanwa0.7 (weighted, asymmetric)


Maximal Enrichment reached 9.333333


Number of similar hits 6


Number of target hits 0


Threshold 1.6554017





Optimizing parameters for descriptor PF, metric Euclideanan0.7 (asymmetric, normalized)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Optimizing parameters for descriptor PF, metric Euclideanwan0.7 (weighted, asymmetric, normalized)


Maximal Enrichment reached 9.333333


Number of similar hits 5


Number of target hits 0


Threshold 0.0





Processing finished at Wed Jan 24 13:57:54 CET 2007





HitStatistics - Molecular Descriptor Screening Statistics 3.1.6,


(C) 2002-2005 ChemAxon Ltd.





No metric selected for descriptor PF, using all available metrics.


No metric selected for descriptor PF, using all available metrics.


No metric selected for descriptor CF, using all available metrics.





Processing started at Wed Jan 24 13:57:56 CET 2007


Processing finished at Wed Jan 24 13:58:00 CET 2007


.


. Now the output text file example_statistics.stat


. containing statistics about the performance of screening with the


. parametrized metrics is displayed.


.


Target file name hit-target


Test set file name hit-test


Query file name opt-query





Number of queries 6


Number of test set elements 7


Number of targets 950





Descr Metric Enrichment SelectivityEffectiveness ActiveHitDistribution Threshold Test Hits Target Hits Before actives


PF Tanimoto 1.000 0.500 -1.000 0.200 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclidean 1.000 0.500 -1.000 15.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotot 1.000 0.500 -1.000 0.922 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotos 1.000 0.500 -1.000 0.744 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotoa 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotosa 1.000 0.500 -1.000 0.283 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideant 1.000 0.500 -1.000 104.269 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideann 1.000 0.500 -1.000 0.805 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwn 1.000 0.500 -1.000 0.037 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideana 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwa 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanan 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwan 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimoto 1.000 0.500 -1.000 0.200 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclidean 1.000 0.500 -1.000 15.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotot0.7 1.000 0.500 -1.000 0.902 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotos0.7 1.000 0.500 -1.000 0.794 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotoa0.7 1.000 0.500 -1.000 0.314 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Tanimotosa0.7 1.000 0.500 -1.000 0.143 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideant0.7 1.000 0.500 -1.000 59.224 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideann0.7 1.000 0.500 -1.000 0.754 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwn0.7 1.000 0.500 -1.000 0.051 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideana0.7 1.000 0.500 -1.000 1.655 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwa0.7 1.000 0.500 -1.000 1.655 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanan0.7 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


PF Euclideanwan0.7 1.000 0.500 -1.000 0.000 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Tanimoto 1.000 0.500 -1.000 0.200 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Euclidean 1.000 0.500 -1.000 10.000 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Tanimotot 1.000 0.500 -1.000 0.688 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Tanimotoa 1.000 0.500 -1.000 0.278 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Euclideant 1.000 0.500 -1.000 9.381 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Euclideann 1.000 0.500 -1.000 0.532 7 950 950 950, 0, 0, 0, 0, 0, 0


CF Euclideana 1.000 0.500 -1.000 3.857 7 950 950 950, 0, 0, 0, 0, 0, 0

ChemAxon efa1591b5a

24-01-2007 14:47:15

Hello,





apparently, this is a bug. Try to use the memory mode of hitstatistics by specifying the -f flag in the command-line. You need to edit the appropriate batch (.bat) or shell script file and add the -f flag in the appropriate lines in which hitstatistics is executed.





Sorry for this trouble, we try to fix this bug in a future release of JChem.





Thank you for reporting this problem.





HTH; please get back to this forum in case you still experience any odd behaviour of the program.





Regards,


Miklos