{
  "id": 448430,
  "title": "Hard coded rules for null values in reactivity profile?",
  "url": "/competitions/stanford-ribonanza-rna-folding/discussion/448430",
  "author_name": "",
  "post_date": "2023-10-19T16:35:28.558827800Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Are there any rules (derived from the experimental procedure, etc) that one can hardcode into our ML pipeline to identify the sequence positions with reactivity == null? In other words, given an arbitrary RNA sequence, is there a rule like \"the first N positions and last M positions will always be null\" or something like that?</p>",
  "messages": [
    {
      "id": "2489044",
      "postDate": "10/19/2023 16:35:28",
      "content": "<p>Are there any rules (derived from the experimental procedure, etc) that one can hardcode into our ML pipeline to identify the sequence positions with reactivity == null? In other words, given an arbitrary RNA sequence, is there a rule like \"the first N positions and last M positions will always be null\" or something like that?</p>",
      "rawMarkdown": "Are there any rules (derived from the experimental procedure, etc) that one can hardcode into our ML pipeline to identify the sequence positions with reactivity == null? In other words, given an arbitrary RNA sequence, is there a rule like \"the first N positions and last M positions will always be null\" or something like that?",
      "votes": null
    },
    {
      "id": "2489129",
      "postDate": "10/19/2023 17:42:14",
      "content": "<p>For the current experiments, the nulls at the beginning and end are of a consistent length, however the ask is to still predict data for those positions - you can find more information at this topic: <a href=\"https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122\" target=\"_blank\">https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122</a></p>",
      "rawMarkdown": "For the current experiments, the nulls at the beginning and end are of a consistent length, however the ask is to still predict data for those positions - you can find more information at this topic: https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2489129,
      "author_name": "jonathanromano",
      "author_url": "",
      "post_date": "10/19/2023 17:42:14",
      "content": "<p>For the current experiments, the nulls at the beginning and end are of a consistent length, however the ask is to still predict data for those positions - you can find more information at this topic: <a href=\"https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122\" target=\"_blank\">https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2489044": "Are there any rules (derived from the experimental procedure, etc) that one can hardcode into our ML pipeline to identify the sequence positions with reactivity == null? In other words, given an arbitrary RNA sequence, is there a rule like \"the first N positions and last M positions will always be null\" or something like that?",
    "2489129": "For the current experiments, the nulls at the beginning and end are of a consistent length, however the ask is to still predict data for those positions - you can find more information at this topic: https://www.kaggle.com/competitions/stanford-ribonanza-rna-folding/discussion/441122"
  },
  "source": "meta"
}