Difference between revisions of "Atomtype"

From Open Babel
Jump to: navigation, search
(Explaining some dependencies in expert systems)
(Data (PATTY expert system): types.txt added)
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
The atom typing algorithm is based on a flexible [[PATTY]] expert  system. This assignment requires previously assigned [[Aromaticity]] and [[Hybridisation]] informations.
+
The atom typing algorithm is based on a flexible [[PATTY]] expert  system ([[Article:bs93]]). This assignment requires previously assigned [[Aromaticity]] and [[Hybridisation]] informations. Some basic atom typing rules were e.g. published by ([[Article:ksfams96]])
 
==Data ([[PATTY]] expert system)==
 
==Data ([[PATTY]] expert system)==
* [http://cvs.sourceforge.net/viewcvs.py/openbabel/openbabel/data/atomtyp.txt?view=markup atomtyp.txt]
+
* [http://openbabel.svn.sourceforge.net/viewvc/openbabel/openbabel/trunk/data/atomtyp.txt?view=markup atomtyp.txt]
==References==
+
* [http://openbabel.svn.sourceforge.net/viewvc/openbabel/openbabel/trunk/data/types.txt?view=markup types.txt]
@Article{bs93,
+
  author = {B. L. Bush and R. P. Sheridan},
+
  title = {{PATTY}: {A} {P}rogrammable {A}tom {T}yper and {L}anguage for {A}utomatic
+
          {C}lassification of {A}toms in {M}olecular {D}atabases.},
+
  journal = {J. Chem. Inf. Comput. Sci.},
+
  year = {1993},
+
  volume = {33},
+
  pages = {756-762},
+
}
+
 
+
@Article{ksfams96,
+
  author = {S. K. Kearsley and S. Sallamack and E. M. Fluder and J. D. Andose
+
            and R. T. Mosley and R. P. Sheridan},
+
  title = {{C}hemical {S}imilarity {U}sing {P}hysiochemical {P}roperty {D}escriptors},
+
  journal = {J. Chem. Inf. Comput. Sci.},
+
  year = {1996},
+
  volume = {36},
+
  pages = {118--127},
+
  abstract = {Similarity searches using topological descriptors have proved extremely
+
        useful in aiding large-scale screening. We describe alternative
+
        forms of the atom pair (Carhart et al. J. Chem. Inf. Comput. Sci.
+
        1985, 25,64-73.) and topological torsion (Nilakantan et al. J. Chem.
+
        Inf. Comput. Sci. 1987, 27,82-85.) descriptors that use physiochemical
+
        atom types. These types are based on binding property class, atomic
+
        log P contribution, and partial atomic charges. The new descriptors
+
        are meant to be more fuzzy than the original descriptors. We propose
+
        objective criteria for determining how effective one descriptor
+
        is versus another in selecting active compounds from large databases.
+
        Using these criteria, we run similarity searches over the Derwent
+
        Standard Drug File with ten typical druglike probes. The new descriptors
+
        are not as good as the original descriptors in selecting actives
+
        if one considers the average over all probes, but the new descriptors
+
        do better for several individual probes. Generally we find that
+
        whether one descriptor does better than another varies from probe
+
        to probe in a way almost impossible to predict a priori. Most importantly,
+
        we find that different descriptors typically select very different
+
        sets of actives. Thus it is advantageous to run similarity probes
+
        with several types of descriptors.},
+
  doi = {[http://dx.doi.org/10.1021/ci950274j 10.1021/ci950274j]}
+
}
+

Latest revision as of 12:41, 1 December 2009

The atom typing algorithm is based on a flexible PATTY expert system (Article:bs93). This assignment requires previously assigned Aromaticity and Hybridisation informations. Some basic atom typing rules were e.g. published by (Article:ksfams96)

Data (PATTY expert system)