ML.CompositeRun.CompositeRun:
class to store parameters for and results from Composite building
This class has a default set of fields which are added to the database.
ML.Descriptors.CompoundDescriptors.CompoundDescriptorCalculator:
used for calculating descriptors
This is the central point for descriptor calculation
**Notes**
- There are two kinds of descriptors this cares about:
1) *Simple Descriptors* can be calculated solely using atomic descriptor
values and the composition of the compound.
ML.files.ReFile:
convenience class for dealing with files with comments
blank (all whitespace) lines, and lines beginning with comment
characters are skipped.
Chem.FastSDMolSupplier.FastSDMolSupplier:
A wrapper around an SDMolSupplier that precomputes and stores
molecular indices (via text processing) to allow quick length
calculations and random access.
ML.Cluster.Clusters.Cluster:
a class for storing clusters/data
**General Remarks**
- It is assumed that the bottom of any cluster hierarchy tree is composed of
the individual data points which were clustered.
Chem.Fingerprints.SimilarityScreener.ThresholdScreener:
Used to return all compounds that have a similarity
to the probe beyond a threshold value
**Notes**:
- This is as lazy as possible, so the data source isn't
queried until the client asks for a hit.
ML.Neural.Trainers.BackProp:
implement back propagation (algorithm on pp 153-154 of Patterson)
I don't *think* that I've made any assumptions about the connectivity of
the net (i.e.
ML.InfoTheory.rdInfoTheory.InfoBitRanker:
A class to rank the bits from a series of labelled fingerprints
A simple demonstration may help clarify what this class does.
Geometry.rdGeometry.Point3D:
A class to represent a three-dimensional point
The x, y, and z coordinates can be read and written using either attributes
(i.e.
Chem.Pharm2D.Signature.Pharm2DSig:
**Notes on Use**
- After any changes, the Init() method should be called
**Important Attributes**
- patterns: a list of SmartsPatterns used to determine the features
patterns can be initialized directly from the list or from a list of SMARTS
strings
- bins: the list of distance bins to be used.
Dbase.DbConnection.DbConnect:
This class is intended to abstract away many of the details of
interacting with databases.
Chem.Pharm2D.LazyGenerator.Generator:
Important attributes:
- mol: the molecules whose signature is being worked with
- sig: the signature to be used to decide what bits are set in the
molecule.
VLib.Filter.FilterNode:
base class for nodes which filter their input
Assumptions:
- filter function takes a number of arguments equal to the
number of inputs we have.
Chem.Fingerprints.DbFpSupplier.RandomAccessDbFpSupplier:
DbFp supplier supporting random access:
>>> import os.path
>>> from Dbase.DbConnection import DbConnect
>>> fName = RDConfig.RDTestDatabase
>>> conn = DbConnect(fName,'simple_combined')
>>> suppl = RandomAccessDbFpSupplier(conn.GetData())
>>> len(suppl)
12
we can pull individual fingerprints:
>>> fp = suppl[5]
>>> fp.GetNumBits()
128
>>> fp.GetNumOnBits()
54
a standard loop over the fingerprints:
>>> fps = []
>>> for fp in suppl:
...
Chem.Fingerprints.DbFpSupplier.ForwardDbFpSupplier:
DbFp supplier supporting only forward iteration
>>> import os.path
>>> from Dbase.DbConnection import DbConnect
>>> fName = RDConfig.RDTestDatabase
>>> conn = DbConnect(fName,'simple_combined')
>>> suppl = ForwardDbFpSupplier(conn.GetData())
we can loop over the supplied fingerprints:
>>> fps = []
>>> for fp in suppl:
...
VLib.Transform.TransformNode:
base class for nodes which filter their input
Assumptions:
- transform function takes a number of arguments equal to the
number of inputs we have.
Chem.Pharm2D.SigFactory.SigFactory:
SigFactory's are used by creating one, setting the relevant
parameters, then calling the GetSignature() method each time a
signature is required.
ML.Composite.Composite.Composite:
a composite model
**Notes**
- adding a model which is already present just results in its count
field being incremented and the errors being averaged.
ML.Composite.BayesComposite.BayesComposite:
a composite model using Bayesian statistics in the Decision Proxy
**Notes**
- typical usage:
1) grow the composite with AddModel until happy with it
2) call AverageErrors to calculate the average error values
3) call SortModels to put things in order by either error or count
4) call Train to update the Bayesian stats.
SimDivFilters.SimilarityPickers.TopNOverallPicker:
A class for picking the top N overall best matches across a library
Connect to a database and build molecules:
>>> import Chem
>>> import os.path
>>> from Dbase.DbConnection import DbConnect
>>> dbName = RDConfig.RDTestDatabase
>>> conn = DbConnect(dbName,'simple_mols1')
>>> [x.upper() for x in conn.GetColumnNames()]
['SMILES', 'ID']
>>> mols = []
>>> for smi,id in conn.GetData():
...
SimDivFilters.SimilarityPickers.SpreadPicker:
A class for picking the best matches across a library
Connect to a database:
>>> import Chem
>>> import os.path
>>> from Dbase.DbConnection import DbConnect
>>> dbName = RDConfig.RDTestDatabase
>>> conn = DbConnect(dbName,'simple_mols1')
>>> [x.upper() for x in conn.GetColumnNames()]
['SMILES', 'ID']
>>> mols = []
>>> for smi,id in conn.GetData():
...
ML.Data.MLData.MLDataSet:
A data set for holding general data (floats, ints, and strings)
**Note**
this is intended to be a read-only data structure
(i.e.
ML.Data.MLData.MLQuantDataSet:
a data set for holding quantized data
**Note**
this is intended to be a read-only data structure
(i.e.