{
  "id": 492816,
  "title": "Purpose and object of prediction",
  "url": "/competitions/leash-BELKA/discussion/492816",
  "author_name": "",
  "post_date": "2024-04-11T03:53:31.382185900Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I still don't get the objective. <br>\nMy understanding is, In each row, we have 1 smile molecule and 3 smile building blocks and 1 protein and 1 bind result.<br>\nThis rows are repeated for the 3 individual protein.<br>\nLooks like check if a new smile molecule bind with each 3 protein based on 3 building blocks…Am I Right?</p>\n<p>What are the relationship with a new simile molecule and building blocks and protein?<br>\nCan you advise?</p>",
  "messages": [
    {
      "id": "2746110",
      "postDate": "04/11/2024 03:53:31",
      "content": "<p>I still don't get the objective. <br>\nMy understanding is, In each row, we have 1 smile molecule and 3 smile building blocks and 1 protein and 1 bind result.<br>\nThis rows are repeated for the 3 individual protein.<br>\nLooks like check if a new smile molecule bind with each 3 protein based on 3 building blocks…Am I Right?</p>\n<p>What are the relationship with a new simile molecule and building blocks and protein?<br>\nCan you advise?</p>",
      "rawMarkdown": "I still don't get the objective. \nMy understanding is, In each row, we have 1 smile molecule and 3 smile building blocks and 1 protein and 1 bind result.\nThis rows are repeated for the 3 individual protein.\nLooks like check if a new smile molecule bind with each 3 protein based on 3 building blocks...Am I Right?\n\nWhat are the relationship with a new simile molecule and building blocks and protein?\nCan you advise?",
      "votes": null
    },
    {
      "id": "2749070",
      "postDate": "04/12/2024 20:05:52",
      "content": "<p>The building blocks are used to to make the molecule. The fact we have them separated out is for our convenience for building train/test splits, or perhaps using the building blocks directly as a feature. However, the goal is to determine if the final molecule does or does not bind to one of the three targets of interest. Please let me know if that is not clear :)</p>",
      "rawMarkdown": "The building blocks are used to to make the molecule. The fact we have them separated out is for our convenience for building train/test splits, or perhaps using the building blocks directly as a feature. However, the goal is to determine if the final molecule does or does not bind to one of the three targets of interest. Please let me know if that is not clear :)",
      "votes": null
    },
    {
      "id": "2760884",
      "postDate": "04/19/2024 14:54:53",
      "content": "<p>If I do not mistake then you want to mean that, the column named 'molecule_smiles\" contains the final molecule which needs to be checked whether it binds or not with the target protein. And in order to do this, I can use \"buildingblock1_smiles\", \"buildingblock2_smiles\", \"buildingblock3_smiles\" columns as a feature and if necessary can add extra feature with my domain knowledge to predict accurately. <br>\nIf I am wrong please let me know. Thanks in advance.</p>",
      "rawMarkdown": "If I do not mistake then you want to mean that, the column named 'molecule_smiles\" contains the final molecule which needs to be checked whether it binds or not with the target protein. And in order to do this, I can use \"buildingblock1_smiles\", \"buildingblock2_smiles\", \"buildingblock3_smiles\" columns as a feature and if necessary can add extra feature with my domain knowledge to predict accurately. \nIf I am wrong please let me know. Thanks in advance.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2749070,
      "author_name": "chemdatafarmer",
      "author_url": "",
      "post_date": "04/12/2024 20:05:52",
      "content": "<p>The building blocks are used to to make the molecule. The fact we have them separated out is for our convenience for building train/test splits, or perhaps using the building blocks directly as a feature. However, the goal is to determine if the final molecule does or does not bind to one of the three targets of interest. Please let me know if that is not clear :)</p>",
      "votes": null,
      "replies": [
        {
          "id": 2760884,
          "author_name": "tdata885",
          "author_url": "",
          "post_date": "04/19/2024 14:54:53",
          "content": "<p>If I do not mistake then you want to mean that, the column named 'molecule_smiles\" contains the final molecule which needs to be checked whether it binds or not with the target protein. And in order to do this, I can use \"buildingblock1_smiles\", \"buildingblock2_smiles\", \"buildingblock3_smiles\" columns as a feature and if necessary can add extra feature with my domain knowledge to predict accurately. <br>\nIf I am wrong please let me know. Thanks in advance.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2746110": "I still don't get the objective. \nMy understanding is, In each row, we have 1 smile molecule and 3 smile building blocks and 1 protein and 1 bind result.\nThis rows are repeated for the 3 individual protein.\nLooks like check if a new smile molecule bind with each 3 protein based on 3 building blocks...Am I Right?\n\nWhat are the relationship with a new simile molecule and building blocks and protein?\nCan you advise?",
    "2749070": "The building blocks are used to to make the molecule. The fact we have them separated out is for our convenience for building train/test splits, or perhaps using the building blocks directly as a feature. However, the goal is to determine if the final molecule does or does not bind to one of the three targets of interest. Please let me know if that is not clear :)",
    "2760884": "If I do not mistake then you want to mean that, the column named 'molecule_smiles\" contains the final molecule which needs to be checked whether it binds or not with the target protein. And in order to do this, I can use \"buildingblock1_smiles\", \"buildingblock2_smiles\", \"buildingblock3_smiles\" columns as a feature and if necessary can add extra feature with my domain knowledge to predict accurately. \nIf I am wrong please let me know. Thanks in advance."
  },
  "source": "meta"
}