Ouroboros-PropModeling

Quantitative Structure-Activity Relationship (QSAR) modeling and molecular property prediction are well-established technologies that have been widely applied to drug screening and design. In Ouroboros, property prediction models can be directly utilized for **virtual screening or _de novo_ molecular design**. You may upload a CSV file containing compound SMILES and corresponding labels to integrate your own custom predictors into the Ouroboros system (your data and models will remain private and inaccessible to other users).

Click here for example output

1. Please copy and paste your csv file content inside the box: Click for an example csv file input

Or you can upload your csv file. Click for an example csv file

2. Please input your label that indicates the column of smiles in your csv input. In the example above, the label is "smiles": Click for an example smiles label input

3. Please input your label that indicates the column of identity in your csv input. In the example above, the label is "label": Click for an example query label input

4. Select the model you want to use. The default is M2:

M2
M1

5. The test ratio is the ratio of the test set. The training set ratio is 1 - test ratio - validate ratio.
Please input your test ratio here. The ratio should be between 0 and 1: Click for an example test ratio input

6. The validate ratio is the ratio of the validate set. The training set ratio is 1 - test ratio - validate ratio.
Please input your validate ratio here. The ratio should be between 0 and 1: Click for an example validate ratio input

7. The dividing method is the method used to divide the dataset into training set, test set and validate set.
Select the dividing method you want to use. The default is Skeleton:

Skeleton
CombineFP
MACCS
ECFP4:AtomPairs

8. The metric is the metric used to evaluate the performance of the model.
Select the Metric you want to use. The default is SPEARMANR:

AUROC
SPEARMANR

9. Please input your email address here. Your result will be sent to this address.
You will receive an ID for the output model. This ID can be later used on other ouroboros jobs:

(optional) Give your job an ID if you want:

Ouroboros-PropModeling

1. About Ouroboros-PropModeling

2. Submission form