{
  "id": 348761,
  "title": "One woman and three men?",
  "url": "/competitions/open-problems-multimodal/discussion/348761",
  "author_name": "AmbrosM",
  "post_date": "2022-08-29T19:58:29.658000",
  "votes": 79,
  "comment_count": 6,
  "views": 0,
  "content": "<p>The Multiome input data consists of chromosome accessibility and the names of the chromosomes. If we plot a histogram of the Y chromosome accessibility (excluding the zero values) for every donor, we get the following diagrams:</p>\n<p><img src=\"https://i.imgur.com/31Bvt5F.png\" alt=\"sex.png\"></p>\n<p>The histograms of the Y chromosomes illustrate the diversity of the donors: Donor 13176 seems to have (almost) no Y chromosome - maybe the few nonzero values are measuring errors. </p>\n<p>We now can ask (and perhaps some biologist will tell us the answer):</p>\n<ul>\n<li>Do these diagrams tell us that donor 13176 is a woman and the three others are men?</li>\n<li>If we assume that the true values of donor 13176 are all zero, what do these data tell us about the magnitude of the measurement errors?</li>\n<li>Should we create a new (binary) feature for the sex of the donors?</li>\n<li>As the donors are so different, how can we make our models robust to these differences?</li>\n</ul>\n<p>Source code is <a href=\"https://www.kaggle.com/code/ambrosm/msci-eda-which-makes-sense\" target=\"_blank\">here</a>.</p>",
  "messages": [
    {
      "id": 1918742,
      "postDate": "2022-08-29T19:58:29.660Z",
      "content": "<p>The Multiome input data consists of chromosome accessibility and the names of the chromosomes. If we plot a histogram of the Y chromosome accessibility (excluding the zero values) for every donor, we get the following diagrams:</p>\n<p><img src=\"https://i.imgur.com/31Bvt5F.png\" alt=\"sex.png\"></p>\n<p>The histograms of the Y chromosomes illustrate the diversity of the donors: Donor 13176 seems to have (almost) no Y chromosome - maybe the few nonzero values are measuring errors. </p>\n<p>We now can ask (and perhaps some biologist will tell us the answer):</p>\n<ul>\n<li>Do these diagrams tell us that donor 13176 is a woman and the three others are men?</li>\n<li>If we assume that the true values of donor 13176 are all zero, what do these data tell us about the magnitude of the measurement errors?</li>\n<li>Should we create a new (binary) feature for the sex of the donors?</li>\n<li>As the donors are so different, how can we make our models robust to these differences?</li>\n</ul>\n<p>Source code is <a href=\"https://www.kaggle.com/code/ambrosm/msci-eda-which-makes-sense\" target=\"_blank\">here</a>.</p>",
      "rawMarkdown": "The Multiome input data consists of chromosome accessibility and the names of the chromosomes. If we plot a histogram of the Y chromosome accessibility (excluding the zero values) for every donor, we get the following diagrams:\n\n![sex.png](https://i.imgur.com/31Bvt5F.png)\n\nThe histograms of the Y chromosomes illustrate the diversity of the donors: Donor 13176 seems to have (almost) no Y chromosome - maybe the few nonzero values are measuring errors. \n\nWe now can ask (and perhaps some biologist will tell us the answer):\n\n- Do these diagrams tell us that donor 13176 is a woman and the three others are men?\n- If we assume that the true values of donor 13176 are all zero, what do these data tell us about the magnitude of the measurement errors?\n- Should we create a new (binary) feature for the sex of the donors?\n- As the donors are so different, how can we make our models robust to these differences?\n\nSource code is [here](https://www.kaggle.com/code/ambrosm/msci-eda-which-makes-sense).",
      "votes": 77
    },
    {
      "id": 1919215,
      "postDate": "2022-08-30T08:07:45.393Z",
      "content": "<p>Fantastic and extremely useful insights!! Keep up the great work!!</p>",
      "rawMarkdown": "Fantastic and extremely useful insights!! Keep up the great work!!",
      "votes": 1
    },
    {
      "id": 1918794,
      "postDate": "2022-08-29T21:02:09.627Z",
      "content": "<p>Nice insight！</p>",
      "rawMarkdown": "Nice insight！",
      "votes": 1
    },
    {
      "id": 1918769,
      "postDate": "2022-08-29T20:32:34.323Z",
      "content": "<p>WOW ! COOL ! </p>",
      "rawMarkdown": "WOW ! COOL ! ",
      "votes": 1
    },
    {
      "id": 2017155,
      "postDate": "2022-11-04T14:40:32.397Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 1935155,
      "postDate": "2022-09-11T22:07:15.640Z",
      "content": "<p>Thanks for sharing!</p>",
      "rawMarkdown": "Thanks for sharing!",
      "votes": 1
    },
    {
      "id": 1920259,
      "postDate": "2022-08-31T03:34:23.043Z",
      "content": "<p>Thanks for sharing. Good job.</p>",
      "rawMarkdown": "Thanks for sharing. Good job.",
      "votes": 1
    }
  ],
  "comments": [
    {
      "id": 1919215,
      "author_name": "Ravi Ramakrishnan",
      "author_url": "",
      "post_date": "2022-08-30T08:07:45.393000",
      "content": "<p>Fantastic and extremely useful insights!! Keep up the great work!!</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1918794,
      "author_name": "Pengnan Chee",
      "author_url": "",
      "post_date": "2022-08-29T21:02:09.627000",
      "content": "<p>Nice insight！</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1918769,
      "author_name": "Alexander Chervov",
      "author_url": "",
      "post_date": "2022-08-29T20:32:34.323000",
      "content": "<p>WOW ! COOL ! </p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 2017155,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-11-04T14:40:32.397000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1935155,
      "author_name": "Autumn",
      "author_url": "",
      "post_date": "2022-09-11T22:07:15.640000",
      "content": "<p>Thanks for sharing!</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1920259,
      "author_name": "Correlation",
      "author_url": "",
      "post_date": "2022-08-31T03:34:23.043000",
      "content": "<p>Thanks for sharing. Good job.</p>",
      "votes": 1,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1918742": "The Multiome input data consists of chromosome accessibility and the names of the chromosomes. If we plot a histogram of the Y chromosome accessibility (excluding the zero values) for every donor, we get the following diagrams:\n\n![sex.png](https://i.imgur.com/31Bvt5F.png)\n\nThe histograms of the Y chromosomes illustrate the diversity of the donors: Donor 13176 seems to have (almost) no Y chromosome - maybe the few nonzero values are measuring errors. \n\nWe now can ask (and perhaps some biologist will tell us the answer):\n\n- Do these diagrams tell us that donor 13176 is a woman and the three others are men?\n- If we assume that the true values of donor 13176 are all zero, what do these data tell us about the magnitude of the measurement errors?\n- Should we create a new (binary) feature for the sex of the donors?\n- As the donors are so different, how can we make our models robust to these differences?\n\nSource code is [here](https://www.kaggle.com/code/ambrosm/msci-eda-which-makes-sense).",
    "1919215": "Fantastic and extremely useful insights!! Keep up the great work!!",
    "1918794": "Nice insight！",
    "1918769": "WOW ! COOL ! ",
    "2017155": "",
    "1935155": "Thanks for sharing!",
    "1920259": "Thanks for sharing. Good job."
  }
}