helma / opentox-model

OpenTox lazar
=============

* An OpenTox REST Webservice (http://www.opentox.org)
* Ruby implementation of the lazar algorithm
* Implements a subset of the OpenTox algorithm/model API (http://opentox.org/wiki/opentox/Structure).
* Current version runs at http://webservices.in-silico.ch/lazar/v0

REST API

Description             Method  URI               Parameters      Result                Status codes

Get index of all        GET     /models                           List of model URIs    200, 500
models  

Get model description   GET     /model{id}                        YAML with model       200, 202, 404, 500
                                                                  representation  

Create a prediction     POST    /                 dataset_uri     Model URI             200, 500
model                                             

Delete a model          DELETE  /{id}                                                   200, 404, 500

Create a prediction     POST    /{id}             compound_uri    Prediction URI        200, 404, 500

Get index of all        GET     /predictions                      List of prediction    200, 500
predictions                                                       URIs            

Show a prediction       GET     /prediction/{id}                  YAML with prediction  200, 202, 404, 500
                                                                  result    

Show neighbors for a    GET     /prediction/{id}                  YAML with neighbors   200, 404, 500
prediction                      /neighbors

Show relevant features  GET     /prediction/{id}                  YAML with relevant    200, 404, 500
for a prediction                /features                         features

Delete a prediction     DELETE  /prediction/{id}                                        200, 404, 500

EXAMPLES

Model creation:

  Create a dataset from a csv file:

    curl -X POST -F name="Hamster carcinogenicity" -F file=@hamster_carcinogenicity.csv  
    http://webservices.in-silico.ch/dataset/v0/

    returns a {dataset_uri}

  Create a prediction model from a dataset:

    curl -X POST -d dataset_uri={dataset_uri} http://webservices.in-silico.ch/lazar/v0/models

    returns a {model_uri}

  Get model information:

    curl {model_uri} (returns status code 202, if model is still under construction)

Prediction:

  Create a compound URI:

    curl -X POST -d name=Phenylhydrazine http://webservices.in-silico.ch/compound/v0/

    returns http://webservices.in-silico.ch/compound/v0/NNc1ccccc1

  Predict the compound:

    curl -X POST -d compound_uri=http://webservices.in-silico.ch/compound/v0/NNc1ccccc1 {model_uri}

    returns a {prediction_uri}

  Show the prediction:

    curl {prediction_uri} (returns status code 202, if prediction is still running)

  Show prediction neighbors:

    curl {prediction_uri}/neighbors

  Show relevant features for a prediction:

    curl {prediction_uri}/features

INSTALLATION

Run 'rake install' to install the required gems

Copyright (c) 2009 Christoph Helma. See LICENSE for details.