phase_find_common Command Help
Command: $SCHRODINGER/phase_find_common
phase_find_commmon - Finds and scores common pharmacophore hypotheses. The basic
approach involves the use of pharmacophore shape similarity to generate numerous
consensus alignments of a set of actives, followed by analysis of each consensus
alignment to identify pharmacophores that contain the requested number of sites
and that satisfy all the matching requirements.
Usage: phase_find_common <project>.phzip [options]
<project>.phzip - Zipped Phase project containing structures and pharmacophore
sites. See $SCHRODINGER/utilities/phase_project for further
details on creating and archiving projects. All results are
returned in <jobName>_find_common.zip, where <jobName> is the
base name of <project>.phzip. Use -JOB <jobName> to override.
Unzipping <jobName>_find_common.zip will create the directory
<jobName>_find_common with the following files/directories:
find_common.opt - Options supplied to this program.
scores.csv - IDs and scores for the requested number of
pharmacophore hypotheses, grouped by the
number of sites and sorted by decreasing
Survival score.
hypotheses - Directory containing the hypothesis files.
alignments - Directory containing aligned actives and
inactives, and match information for each
hypothesis.
Options: [-JOB <jobName>]
[-HOST <host>[:<n>]]
[-run <runName>]
[-ref <list>]
[-sites <min>:<max>]
[-ex]
[-miss <m> [-fraction]]
[-prefer <pmin>]
[-redun <tol>]
[-keep <maxHypo>]
[-try <maxConfs>]
[-inplace]
[-site <weight>]
[-vect <weight>]
[-vol <weight>]
[-select <weight>]
[-logm <weight>]
[-inactive <weight> | -noinactive]
[-pos <actives>.phzip -neg <decoys>.phzip -alpha <value> | -nobedroc]
[-LOCALDRIVER [-USEDRIVERNODE]]
[-LOCAL]
[-TMPDIR <dir>]
[-WAIT]
[-NICE]
[-verbosity <level>]
[-h|-help]
-JOB <jobName> - Job name. The default is to use the base name of
<project>.phzip.
-HOST <host>[:<n>] - Run job on a remote host. Include \":<n>\" to split over
<n> CPUs. The maximum allowed number of CPUs is equal
to the number of actives, or the number of reference
ligands if using -ref <list>.
-run <runName> - Run name property (e.g., \"run1\") to store in each
hypothesis. If omitted, it will be the original project
path.
-ref <list> - Integer list of actives to treat as reference ligands.
For example,
<list> Reference Ligands
3:5,8,9 mol_3, mol_4, mol_5, mol_8, mol_9
:3,5:7 mol_1, mol_2, mol_3, mol_5, mol_6, mol_7
By default, each active is treated as a reference.
-sites <min>:<max> - Search each reference ligand for common pharamacophores
containing between <min> and <max> sites. The legal
range is 3:7, and the default is 4:6. The actual search
proceeds from <max> down to <min>, and halts before
reaching <min> if common pharmacophores containing more
than <min> sites are found. Use -ex to force the full
range to be considered. This procedure is followed
independently for each reference ligand conformer, so
it is still possible to obtain common pharmacophores
that contain different numbers of sites even if -ex is
not used. Precise rules regarding which sites in a
given ligand must/cannot be matched may be defined by
modifying the project file SiteMask.tab.
-ex - Consider the full range of sites, from <max> to <min>.
-miss <m> - The maximum number of actives/active groups that may be
missed by common pharmacophores of a given reference
ligand. Initally, all actives/active groups are
required to match, but if no solutions are found for a
particular reference ligand, the number of misses is
iteratively increased to 1,2,...,<m>, and halts before
reaching <m> if common pharmacophores are found. By
default, <m> is 1/2 the number of actives/active
groups. An active group, such as the ionic/tautomeric
states of given ligand, may be defined by modifying the
LIGAND_GROUP values in the project file MasterData.tab.
-fraction - Interpret <m> as the fraction of actives/active groups
that may be missed. Accordingly, <m> is multiplied by
the total number of actives/active groups <n>, and then
rounded to the nearest integer. The resulting integer
is reduced to <n> - 2 if it exceeds <n> - 2.
-prefer <pmin> - The preferred minimum number of sites when iteratively
increasing the number of misses. If this option is
used, the number of misses will be increased until
common pharmacophores containing at least <pmin> sites
are found. If the number of misses proceeds all the way
to <m> without yielding pharmacophores that contain at
least <pmin> sites, the reported solutions will contain
the largest number of sites that produced the smallest
number of misses. The rationale for this option is to
promote discovery of common pharmacophores containing
larger numbers of sites, even if it means matching
fewer actives/active groups. This option is off by
default.
-redun <tol> - Site point positional difference for elimination of
redundant pharmacophores. The default is 0.25.
-keep <maxHypo> - Maximum number of hypotheses to retain for each number
of sites. The default is 10 hypotheses with n sites, 10
hypotheses with n-1 sites, etc., where n is the largest
number of sites that produced common pharmacophores.
-try <maxConfs> - When aligning actives to the conformers of a given
reference ligand, consider only the first <maxConfs>
reference conformers, proceeding over those conformers
in order of decreasing average shape similarity to the
other actives. By default, all reference conformers are
considered. This option is normally used only for
pedagogical or debugging purposes, in combination with
-verbosity 3, to see a high level of detail for a small
number of conformers. It does not significantly reduce
the computation time.
-inplace - Perceive and score common pharmacophores using the
supplied poses without doing any alignments. Intended
only for the case where each ligand is represented by a
single conformer and ligands are already aligned in the
same frame of reference.
-site <weight> - Site score weight to use when computing Survival score.
Must be >= 0. The default is 1.
-vect <weight> - Vector score weight. Must be >= 0. The default is 1.
-vol <weight> - Volume score weight. Must be >= 0. The default is 1.
-select <weight> - Selectivity score weight. Must be >= 0. The default is
1.
-logm <weight> - log10(#Matches) weight. Must be >= 0. The default is 1.
-inactive <weight> - Inactive score weight. Relevant only when the project
contains compounds that have been marked as inactive.
Must be >= 0. The default is 1.
-noinactive - Do not perform inactive scoring, even if the project
contains inactives.
-pos <actives>.phzip - Zipped project containing a set of known actives for
validation of common pharmacophore hypotheses. The
actives in <actives>.phzip and the inactives in
<decoys>.phzip are screened against each surviving
hypothesis to yield a BEDROC score on the interval
[0, 1], where higher values indicate a superior ability
to distinguish actives from decoys. For details, see
J. Chem. Inf. Model. 2007, 47, 488-508. If this option
is omitted, the actives will be the same ones used to
identify common pharmacophores, but they will be
represented with default phase_database conformers. To
screen the same conformers that were used for common
pharmacophore identification, supply <project>.phzip as
the actives project.
-neg <decoys>.phzip - Zipped project containing a set of inactives to use as
decoys. This may also be <project>.phzip, as long as
that project contains inactives. By default, a set of
1000 diverse decoys in the Phase installation are used.
-alpha <value> - BEDROC early recognition parameter. Higher values
increase the importance of finding actives early in a
virtual screen. The default is 20.
-nobedroc - Do not compute BEDROC scores.
-LOCALDRIVER - Run driver process on the local host. Subjobs will run
on whichever host was requested by -HOST.
-USEDRIVERNODE - Keep one subjob running on the driver node. This is
desirable when a -LOCALDRIVER job is being run as part
of a workflow where the local host is a queued node. It
may not be desirable if the local host is a desktop
machine with limited computing power. This flag is
always on when -LOCALDRIVER is not specified.
-LOCAL - Store temporary job files in current directory.
-TMPDIR <dir> - Store temporary job files in <dir>.
-WAIT - Do not return until job completes.
-NICE - Run job at reduced priority.
-verbosity <level> - Verbosity of output to log file:
1 = low
2 = medium (default)
3 = high
High verbosity is discouraged unless the number of
reference conformers being considered is small (see
-try <maxConfs>).
-h|-help - Show this help message and exit.