{
  "id": 491229,
  "title": "We must predict \"binds to any of the three protein targets\"?",
  "url": "/competitions/leash-BELKA/discussion/491229",
  "author_name": "",
  "post_date": "2024-04-05T04:46:23.400581200Z",
  "votes": 3,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hello,</p>\n<p>For the test set, what are we supposed to predict?</p>\n<p>The training set has three protein targets.</p>\n<p>What are you interested into:</p>\n<ul>\n<li>probability of binding to any of the three proteins?</li>\n<li>probability of binding to all three proteins?</li>\n</ul>",
  "messages": [
    {
      "id": "2736184",
      "postDate": "04/05/2024 04:46:23",
      "content": "<p>Hello,</p>\n<p>For the test set, what are we supposed to predict?</p>\n<p>The training set has three protein targets.</p>\n<p>What are you interested into:</p>\n<ul>\n<li>probability of binding to any of the three proteins?</li>\n<li>probability of binding to all three proteins?</li>\n</ul>",
      "rawMarkdown": "Hello,\n\nFor the test set, what are we supposed to predict?\n\nThe training set has three protein targets.\n\nWhat are you interested into:\n- probability of binding to any of the three proteins?\n- probability of binding to all three proteins?",
      "votes": null
    },
    {
      "id": "2736211",
      "postDate": "04/05/2024 05:06:02",
      "content": "<p>Ok, I had a look at the test file. For each test molecule, you specify which target you are interested into.</p>",
      "rawMarkdown": "Ok, I had a look at the test file. For each test molecule, you specify which target you are interested into.",
      "votes": null
    },
    {
      "id": "2736408",
      "postDate": "04/05/2024 07:17:39",
      "content": "<p><a href=\"https://www.kaggle.com/francoisberenger\" target=\"_blank\">@francoisberenger</a> probability of binding to the target</p>",
      "rawMarkdown": "francoisberenger probability of binding to the target",
      "votes": null
    },
    {
      "id": "2738361",
      "postDate": "04/06/2024 09:41:15",
      "content": "<p>The output should be a binary number (0 or 1)</p>",
      "rawMarkdown": "The output should be a binary number (0 or 1)",
      "votes": null
    },
    {
      "id": "2738368",
      "postDate": "04/06/2024 09:46:39",
      "content": "<p>No. <a href=\"https://www.kaggle.com/competitions/leash-BELKA/overview/evaluation\" target=\"_blank\">See the example</a>.</p>",
      "rawMarkdown": "No. [See the example](https://www.kaggle.com/competitions/leash-BELKA/overview/evaluation).",
      "votes": null
    },
    {
      "id": "2740925",
      "postDate": "04/08/2024 02:39:52",
      "content": "<p>So,we should predict the drug–target binding affinity，and the number of the output is binding affinity score?</p>",
      "rawMarkdown": "So,we should predict the drug–target binding affinity，and the number of the output is binding affinity score?",
      "votes": null
    },
    {
      "id": "2752078",
      "postDate": "04/14/2024 17:37:36",
      "content": "<p>test.csv appears that each molecule is represented by three rows, corresponding to three different proteins (targets), each with a unique ID. Therefore, I assume that when creating the submission file, each ID's prediction corresponds to the respective molecule and target combination.</p>",
      "rawMarkdown": "test.csv appears that each molecule is represented by three rows, corresponding to three different proteins (targets), each with a unique ID. Therefore, I assume that when creating the submission file, each ID's prediction corresponds to the respective molecule and target combination.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2736211,
      "author_name": "francoisberenger",
      "author_url": "",
      "post_date": "04/05/2024 05:06:02",
      "content": "<p>Ok, I had a look at the test file. For each test molecule, you specify which target you are interested into.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2736408,
          "author_name": "seshurajup",
          "author_url": "",
          "post_date": "04/05/2024 07:17:39",
          "content": "<p><a href=\"https://www.kaggle.com/francoisberenger\" target=\"_blank\">@francoisberenger</a> probability of binding to the target</p>",
          "votes": null,
          "replies": [
            {
              "id": 2738361,
              "author_name": "radleighdeveloper",
              "author_url": "",
              "post_date": "04/06/2024 09:41:15",
              "content": "<p>The output should be a binary number (0 or 1)</p>",
              "votes": null,
              "replies": [
                {
                  "id": 2738368,
                  "author_name": "shlomoron",
                  "author_url": "",
                  "post_date": "04/06/2024 09:46:39",
                  "content": "<p>No. <a href=\"https://www.kaggle.com/competitions/leash-BELKA/overview/evaluation\" target=\"_blank\">See the example</a>.</p>",
                  "votes": null,
                  "replies": [
                    {
                      "id": 2740925,
                      "author_name": "qwerrtyty",
                      "author_url": "",
                      "post_date": "04/08/2024 02:39:52",
                      "content": "<p>So,we should predict the drug–target binding affinity，and the number of the output is binding affinity score?</p>",
                      "votes": null,
                      "replies": []
                    }
                  ]
                }
              ]
            }
          ]
        }
      ]
    },
    {
      "id": 2752078,
      "author_name": "narareddy",
      "author_url": "",
      "post_date": "04/14/2024 17:37:36",
      "content": "<p>test.csv appears that each molecule is represented by three rows, corresponding to three different proteins (targets), each with a unique ID. Therefore, I assume that when creating the submission file, each ID's prediction corresponds to the respective molecule and target combination.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2736184": "Hello,\n\nFor the test set, what are we supposed to predict?\n\nThe training set has three protein targets.\n\nWhat are you interested into:\n- probability of binding to any of the three proteins?\n- probability of binding to all three proteins?",
    "2736211": "Ok, I had a look at the test file. For each test molecule, you specify which target you are interested into.",
    "2736408": "francoisberenger probability of binding to the target",
    "2738361": "The output should be a binary number (0 or 1)",
    "2738368": "No. [See the example](https://www.kaggle.com/competitions/leash-BELKA/overview/evaluation).",
    "2740925": "So,we should predict the drug–target binding affinity，and the number of the output is binding affinity score?",
    "2752078": "test.csv appears that each molecule is represented by three rows, corresponding to three different proteins (targets), each with a unique ID. Therefore, I assume that when creating the submission file, each ID's prediction corresponds to the respective molecule and target combination."
  },
  "source": "meta"
}