{
  "id": 571029,
  "title": "coordinate origin",
  "url": "/competitions/stanford-rna-3d-folding/discussion/571029",
  "author_name": "",
  "post_date": "2025-04-01T04:02:20.803158300Z",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>how the values of x, y and z what is assumed to be the origin for calculations</p>",
  "messages": [
    {
      "id": "3166818",
      "postDate": "04/01/2025 04:02:20",
      "content": "<p>how the values of x, y and z what is assumed to be the origin for calculations</p>",
      "rawMarkdown": "how the values of x, y and z what is assumed to be the origin for calculations",
      "votes": null
    },
    {
      "id": "3167190",
      "postDate": "04/01/2025 10:44:04",
      "content": "<p>Here is the code for conversion from pdb files of simulated RNA to x, y, z formatted as Stanford RNA 3D Folding competition: <a href=\"https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset\" target=\"_blank\">https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset</a></p>\n<p>However, I'm not sure if they translate the structure in pdb files before converting to origin O(0,0,0) or not.</p>",
      "rawMarkdown": "Here is the code for conversion from pdb files of simulated RNA to x, y, z formatted as Stanford RNA 3D Folding competition: https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset\n\nHowever, I'm not sure if they translate the structure in pdb files before converting to origin O(0,0,0) or not.",
      "votes": null
    },
    {
      "id": "3167793",
      "postDate": "04/01/2025 20:39:19",
      "content": "<p>There is no absolute coordinate system, as predictions will be rotated and translated in space to match target coordinates the best way. That means there is no need to have anything placed at a (0,0,0) point.</p>",
      "rawMarkdown": "There is no absolute coordinate system, as predictions will be rotated and translated in space to match target coordinates the best way. That means there is no need to have anything placed at a (0,0,0) point.",
      "votes": null
    },
    {
      "id": "3168405",
      "postDate": "04/02/2025 11:54:55",
      "content": "<p>And the host shared their preprocessing pipeline: <a href=\"https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085\" target=\"_blank\">https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085</a></p>",
      "rawMarkdown": "And the host shared their preprocessing pipeline: https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085",
      "votes": null
    },
    {
      "id": "3173774",
      "postDate": "04/08/2025 10:12:19",
      "content": "<p>here's an example to understand it intuitively:</p>\n<p>Imagine two structures:</p>\n<p><strong>Ground Truth</strong>:</p>\n<pre><code>' atom : (, , )\n' atom : (, , )\n' atom : (, , )\n</code></pre>\n<p><strong>Prediction</strong>:</p>\n<pre><code>' atom : (, , )\n' atom : (, , )\n' atom : (, , )\n</code></pre>\n<p>These structures are identical in terms of relative positions, just translated by <code>(100, 100, 100)</code>. After alignment, they would perfectly match.</p>",
      "rawMarkdown": "here's an example to understand it intuitively:\n\nImagine two structures:\n\n**Ground Truth**:\n\n```\nC1' atom 1: (10, 10, 10)\nC1' atom 2: (12, 10, 10)\nC1' atom 3: (11, 12, 10)\n```\n\n**Prediction**:\n```\nC1' atom 1: (110, 110, 110)\nC1' atom 2: (112, 110, 110)\nC1' atom 3: (111, 112, 110)\n```\n\nThese structures are identical in terms of relative positions, just translated by `(100, 100, 100)`. After alignment, they would perfectly match.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3167190,
      "author_name": "nguyenhoa",
      "author_url": "",
      "post_date": "04/01/2025 10:44:04",
      "content": "<p>Here is the code for conversion from pdb files of simulated RNA to x, y, z formatted as Stanford RNA 3D Folding competition: <a href=\"https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset\" target=\"_blank\">https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset</a></p>\n<p>However, I'm not sure if they translate the structure in pdb files before converting to origin O(0,0,0) or not.</p>",
      "votes": null,
      "replies": [
        {
          "id": 3168405,
          "author_name": "nguyenhoa",
          "author_url": "",
          "post_date": "04/02/2025 11:54:55",
          "content": "<p>And the host shared their preprocessing pipeline: <a href=\"https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085\" target=\"_blank\">https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085</a></p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3167793,
      "author_name": "tilii7",
      "author_url": "",
      "post_date": "04/01/2025 20:39:19",
      "content": "<p>There is no absolute coordinate system, as predictions will be rotated and translated in space to match target coordinates the best way. That means there is no need to have anything placed at a (0,0,0) point.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3173774,
      "author_name": "jaejohn",
      "author_url": "",
      "post_date": "04/08/2025 10:12:19",
      "content": "<p>here's an example to understand it intuitively:</p>\n<p>Imagine two structures:</p>\n<p><strong>Ground Truth</strong>:</p>\n<pre><code>' atom : (, , )\n' atom : (, , )\n' atom : (, , )\n</code></pre>\n<p><strong>Prediction</strong>:</p>\n<pre><code>' atom : (, , )\n' atom : (, , )\n' atom : (, , )\n</code></pre>\n<p>These structures are identical in terms of relative positions, just translated by <code>(100, 100, 100)</code>. After alignment, they would perfectly match.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3166818": "how the values of x, y and z what is assumed to be the origin for calculations",
    "3167190": "Here is the code for conversion from pdb files of simulated RNA to x, y, z formatted as Stanford RNA 3D Folding competition: https://www.kaggle.com/code/tomooinubushi/convert-stanford-ribonanza-rna-folding-dataset\n\nHowever, I'm not sure if they translate the structure in pdb files before converting to origin O(0,0,0) or not.",
    "3167793": "There is no absolute coordinate system, as predictions will be rotated and translated in space to match target coordinates the best way. That means there is no need to have anything placed at a (0,0,0) point.",
    "3168405": "And the host shared their preprocessing pipeline: https://www.kaggle.com/competitions/stanford-rna-3d-folding/discussion/569085",
    "3173774": "here's an example to understand it intuitively:\n\nImagine two structures:\n\n**Ground Truth**:\n\n```\nC1' atom 1: (10, 10, 10)\nC1' atom 2: (12, 10, 10)\nC1' atom 3: (11, 12, 10)\n```\n\n**Prediction**:\n```\nC1' atom 1: (110, 110, 110)\nC1' atom 2: (112, 110, 110)\nC1' atom 3: (111, 112, 110)\n```\n\nThese structures are identical in terms of relative positions, just translated by `(100, 100, 100)`. After alignment, they would perfectly match."
  },
  "source": "meta"
}