Free ChemBioGrid web services hosted at Indiana
A list of current web services freely available from the CICC. In some
cases, the services were developed in conjunction with
organizations outside Indiana University. In a few cases, WSDLs of
others are included.
REST web services
These services can be accessed through REST (simple URL-based
access). To use these services, append the required parameter(s) (for
example, a SMILES string) onto the URL. For example, to get the InChI
Key for Benzene, go to the URL
http://cheminfo.informatics.indiana.edu/rest/format/inchikey/c1ccccc1.
Note that accessing a URL without a parameter will result in an error
Format Conversions:
http://cheminfo.informatics.indiana.edu/rest/format/sdf/ + smiles (SMILES to 2D SD File )
http://cheminfo.informatics.indiana.edu/rest/format/inchi/ + smiles (SMILES to InChI)
http://cheminfo.informatics.indiana.edu/rest/format/inchikey/ + smiles (SMILES to InChI Key)
http://cheminfo.informatics.indiana.edu/rest/db/pubchem/cidinchi/ + inchi (InChI to PubChem CID)
http://cheminfo.informatics.indiana.edu/rest/db/pubchem/cidikey/ + inchikey (InChI Key to PubChem CID)
NIH Tumor Cell Line Predictions:
Predictions for activity in 40 human tumor cell lines, based on Random Forest models using 166 MACCS key descriptors.
http://cheminfo.informatics.indiana.edu/rest/NCI/ + smiles
3D structure in SDF format for a given CID, InChI or InChI key:
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + inchikey (inchikey to 3D)
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + cid (cid to 3D)
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + inchi (inchi to 3D)
2D Descriptors:
http://cheminfo.informatics.indiana.edu/rest/desc/descriptors (To obtain information on all available descriptors)
http://cheminfo.informatics.indiana.edu/rest/desc/descriptors/ + smiles(To calculate descriptor values for a SMILES)
3D structure for a PubChem compound, retrieved from the Pub3D database
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + cid (PubChem CID to 3D SD File)
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + inchi (InChI to 3D SD File)
http://cheminfo.informatics.indiana.edu/rest/db/pub3d/ + inchikey (InChI Key to 3D SD File)
Generation of a 3D structure from a SMILES string, using the smi23d program
http://cheminfo.informatics.indiana.edu/rest/thread/d3.py/SMILES/ + SMILES (2D SMILES to 3D)
3D similarity search on PubChem
http://cheminfo.informatics.indiana.edu/rest/search/threaddb/NHIT/RADIUS/COORDS/
(NHIT is the number of results desired, RADIUS is the search radius and
COORDS is the comma-separated string of 12D shape descriptors)
3DistanceMoments:
Calculates 12-dimensional 3D descriptors for a compound, as used for
the 3D similarity searching, from a PubChem CID:
http://cheminfo.informatics.indiana.edu/rest/3DistanceMoments/ + cid
SOAP web services
These services can be accessed through the WSDL descriptions that are linked, using SOAP.
NIH Tumor Cell Line Predictions.
Predictions for activity in 40 human tumor cell lines, based on Random Forest models using 166 MACCS key descriptors.
http://cheminfov.informatics.indiana.edu:8080/rws/services/ModelDtp?wsdl
Linear Regression:
http://cheminfo.informatics.indiana.edu:8080/rws/services/LinearRegressionModel?
CNN Regression:
http://cheminfo.informatics.indiana.edu:8080/rws/services/CNNRegressionModel?wsdl
RF Regression:
http://cheminfo.informatics.indiana.edu:8080/rws/services/RFRegressionModel?wsdl
LDA Classification:
http://cheminfo.informatics.indiana.edu:8080/rws/services/LDAModel?wsdl
K-means Clustering:
http://cheminfo.informatics.indiana.edu:8080/rws/services/KMeansModel?wsdl
t-Test:
http://cheminfo.informatics.indiana.edu:8080/rws/services/TTest?wsdl
XY Plots:
http://cheminfo.informatics.indiana.edu:8080/rws/services/XYPlots?wsdl
Feature selection:
http://cheminfo.informatics.indiana.edu:8080/rws/services/FeatureSelection?wsdl
Pharmacokinetic Calculator(pkCell):
Calculates pharmacokinetic properties for a structure
http://cheminfo.informatics.indiana.edu:8080/rws/services/PkCell?wsdl
Toxicity Predictions- Random Forest Ensemble:
Ensemble of predictive models for toxicity
http://cheminfo.informatics.indiana.edu:8080/rws/services/ScrippsTox?wsdl
ToxTree:
ToxTree predictions using the Cramer rules, etc.
http://cheminfov.informatics.indiana.edu:8080/tox/services/toxTreeWS?wsdl
Instructions for Linear Regression Models
LinearRegressionModel:
1. Create the model, get model name and some coefficients based on the input data, following is the URL:
http://cheminfo.informatics. indiana.edu/rest/ LinearRegressionModel-
cofficients/which coefficient you want to calculate/The X matrix/The
observed Y vector
there have several coefficients can be calculated:
1) getCoefficients: return The value of the coefficients field
2) getStdErrCoeff: Returns the standard error of the coefficients.
3) getTValueCoeff: Returns the t-value of the coefficients.
4) getPValueCoeff: Returns the p-value of the coefficients.
5) getDfResidual: return The value of the df.residual field
6) getFittedValues: return The value of the fitted.values field
7) getResiduals: return The value of the residuals field
8) buildModel: return the name of the model that was built
The X matrix:
defined as double[] [] in the service, so be sure to put in the url as the following format:
example: [[1, 2], [3, 4], [5, 6], [7, 8]] should be written as 1,2|3,4|5,6|7,8 in the "The X matrix" section.
The observed Y vector:
defined as double[], so plz be sure to write like this: 1,2,3,4 in the "The observed Y vector" section
note: the length of the X matrix should be equal with the length of the Y vector.
2. Get predictions
http://cheminfo.informatics. indiana.edu/rest/ LinearRegressionModel- prediction/modelName/newx/ level/interval
modelName The name of the model
newx The new X data with which to predict, it is defined as double[][]
level At what level should the confidence intervals be returned, it is defined as double
interval Indicates what type of interval should be calculated. Must be
one of <i>confidence</i> or <i>prediction</i>
return A 2D array of doubles with 4 columns containing the
<ol>
<li>predicted values,
<li>upper limit of the interval
<li>lower limit of the interval
<li>standard error of fit
</ol>
throws RWebServiceException if the model is not of type
<code>'lm'</code> or if the columns of the new data are not
of the same number as used to build the model
hrows REXPMismatchException
throws REngineException
throws RSrvException if there is an error in communicating with the Rserve process
Exemplar chembiogrid clients
PubChemSR - a PC desktop client for PubChem
WENDI - An aggregate tool for finding compound-gene and compound-disease relationships using web services
Chem2Bio2RDF - an RDF systems chemical biology / chemogenomics infrastructure