{
  "id": 567304,
  "title": "RNA nucleotide sequence position in 3D",
  "url": "/competitions/stanford-rna-3d-folding/discussion/567304",
  "author_name": "",
  "post_date": "2025-03-09T15:05:59.647743800Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I just curious about the position of the nucleotide in 3D space of this competition. Is there any rule of how researcher make this in 3D? Should we transform the sequence to make the first nucleotide start at the origin x = 0, y = 0, z = 0? Maybe the first nucleotide start with origin then make a model to predict how the next nucleotide should change compare to the previous ?</p>",
  "messages": [
    {
      "id": "3145240",
      "postDate": "03/09/2025 15:05:59",
      "content": "<p>I just curious about the position of the nucleotide in 3D space of this competition. Is there any rule of how researcher make this in 3D? Should we transform the sequence to make the first nucleotide start at the origin x = 0, y = 0, z = 0? Maybe the first nucleotide start with origin then make a model to predict how the next nucleotide should change compare to the previous ?</p>",
      "rawMarkdown": "I just curious about the position of the nucleotide in 3D space of this competition. Is there any rule of how researcher make this in 3D? Should we transform the sequence to make the first nucleotide start at the origin x = 0, y = 0, z = 0? Maybe the first nucleotide start with origin then make a model to predict how the next nucleotide should change compare to the previous ?",
      "votes": null
    },
    {
      "id": "3145305",
      "postDate": "03/09/2025 17:20:36",
      "content": "<p>About submission origin doesn't matter. Predictions will be translated/rotated to best alignment with target. At training is your choice, I think structures are aproximately centered to theyr centroids and not oriented, just as they've been experimentally obtained.</p>",
      "rawMarkdown": "About submission origin doesn't matter. Predictions will be translated/rotated to best alignment with target. At training is your choice, I think structures are aproximately centered to theyr centroids and not oriented, just as they've been experimentally obtained.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3145305,
      "author_name": "sacuscreed",
      "author_url": "",
      "post_date": "03/09/2025 17:20:36",
      "content": "<p>About submission origin doesn't matter. Predictions will be translated/rotated to best alignment with target. At training is your choice, I think structures are aproximately centered to theyr centroids and not oriented, just as they've been experimentally obtained.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3145240": "I just curious about the position of the nucleotide in 3D space of this competition. Is there any rule of how researcher make this in 3D? Should we transform the sequence to make the first nucleotide start at the origin x = 0, y = 0, z = 0? Maybe the first nucleotide start with origin then make a model to predict how the next nucleotide should change compare to the previous ?",
    "3145305": "About submission origin doesn't matter. Predictions will be translated/rotated to best alignment with target. At training is your choice, I think structures are aproximately centered to theyr centroids and not oriented, just as they've been experimentally obtained."
  },
  "source": "meta"
}