{
  "fetched_at": "2026-10-20T08:21:36Z",
  "source_url": "https://www.kaggle.com/competitions/drug-solubility-challenge/data",
  "pages": [
    {
      "name": "rules",
      "content": "- Don't cheat!\n- Ask for help when you need!\n- Apply yourself!\n- Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "<p>ESOL (delaney) is a standard regression dataset containing structures and water solubility data for 1128 compounds. The dataset is widely used to validate machine learning models on estimating solubility directly from molecular structures (as encoded in SMILES strings).</p>\n<p><strong>Reference:</strong></p>\n<p>Delaney, John S. \"ESOL: estimating aqueous solubility directly from molecular structure.\" Journal of chemical information and computer sciences 44.3 (2004): 1000-1005.</p>\n<h2>File descriptions</h2>\n<ul>\n<li><strong>delaney-processed.csv</strong> - the data set for training</li>\n<li><strong>test.csv</strong> - the test set</li>\n</ul>\n<h2>Data fields</h2>\n<p>The delaney-processed data file contains a csv table, in which columns below are used:>\n<ul>\n<li><strong>smiles</strong> - SMILES representation of the molecular structure</li>\n<li><strong>measured log solubility in mols per litre</strong> - Log-scale water solubility of the compound, used as label [log(mol/L)]</li>\n</ul>\n<p>The test data file contains only one column, for which you are required to predict the solubility:\n<ul>\n<li><strong>smiles</strong> - SMILES representation of the molecular structure</li>\n</ul>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "<p>Solubility, the phenomenon of dissolution of solute in solvent to give a homogenous system, is one of the important parameters to achieve desired concentration of drug in systemic circulation for desired pharmacological response. Low aqueous solubility is the major problem encountered with formulation development of new chemical entities as well as for the generic development. More than 40% NCEs (new chemical entities) developed in pharmaceutical industry are practically insoluble in water. </p>\n<p>Solubility is still a major challenge for molecule optimization. Any drug to be absorbed must be present in the form of solution at the site of absorption. The enhancement of the solubility of poorly soluble drugs involves a lot of trials and errors, which calls for AI's guidance. </p>\n\n<h2>Acknowledgements</h2>\n<p>We thank Prof. Jie Zheng and Zhaoping Xiong, Ph.D. for designing this contest.</p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "<p>In our setting, this is a regression problem to predict label values, RMSE is employed as the evaluation metric. The RMSE of a test set is given by:</p>\n<p>\\[ \\text {RMSE} = \\sqrt{ \\sum\\nolimits_{i=1}^{n} \\frac{\\left(y_{predicted,i} - y_{measured,i}  \\right)^2}{n} } \\]</p>\n<h2>Submission Format</h2>\n<p>The file should contain a header and have the following format:</p>\n<pre>smile,LogS<br>O=C1CCCN1,1.68<br></pre>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}