{
  "id": 76414,
  "title": "About 'What am I predicting?'",
  "url": "/competitions/human-protein-atlas-image-classification/discussion/76414",
  "author_name": "",
  "post_date": "2019-01-02T16:34:28.701725800Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Here is the describation in '<strong>What am I predicting</strong>':</p>\n\n<p>All image samples are represented by four filters (stored as individual files), the protein of interest (green) plus three cellular landmarks: nucleus (blue), microtubules (red), endoplasmic reticulum (yellow). The green filter should hence be used to predict the label, and the other filters are used as references.</p>\n\n<p>I am confused about how to use these four pics correctly. What does 'The green filter should hence be used to predict the label, and the other filters are used as references'  mean? Thanks!! \n'</p>",
  "messages": [
    {
      "id": "449100",
      "postDate": "01/02/2019 16:34:28",
      "content": "<p>Here is the describation in '<strong>What am I predicting</strong>':</p>\n\n<p>All image samples are represented by four filters (stored as individual files), the protein of interest (green) plus three cellular landmarks: nucleus (blue), microtubules (red), endoplasmic reticulum (yellow). The green filter should hence be used to predict the label, and the other filters are used as references.</p>\n\n<p>I am confused about how to use these four pics correctly. What does 'The green filter should hence be used to predict the label, and the other filters are used as references'  mean? Thanks!! \n'</p>",
      "rawMarkdown": "Here is the describation in '**What am I predicting**':\n\n All image samples are represented by four filters (stored as individual files), the protein of interest (green) plus three cellular landmarks: nucleus (blue), microtubules (red), endoplasmic reticulum (yellow). The green filter should hence be used to predict the label, and the other filters are used as references.\n\nI am confused about how to use these four pics correctly. What does 'The green filter should hence be used to predict the label, and the other filters are used as references'  mean? Thanks!! \n'",
      "votes": null
    },
    {
      "id": "450113",
      "postDate": "01/04/2019 10:28:29",
      "content": "<p>Afaik the different images stem from different stainings (<a href=\"https://en.wikipedia.org/wiki/Staining\">https://en.wikipedia.org/wiki/Staining</a> ) and therefore different things are visible in the images. All can give you indications on the correct label though.</p>",
      "rawMarkdown": "Afaik the different images stem from different stainings (https://en.wikipedia.org/wiki/Staining ) and therefore different things are visible in the images. All can give you indications on the correct label though.",
      "votes": null
    },
    {
      "id": "450564",
      "postDate": "01/05/2019 09:11:40",
      "content": "<p>Thanks, I know what you mean.Sorry for previous reply.</p>",
      "rawMarkdown": "Thanks, I know what you mean.Sorry for previous reply.",
      "votes": null
    },
    {
      "id": "450672",
      "postDate": "01/05/2019 14:07:58",
      "content": "<p>I'm not taking part in this competition, so you might want to double check my answer in the context of the competition. I have done a good amount of cell staining and <a href=\"https://en.wikipedia.org/wiki/Fluorescence_microscope\">fluorescence microscopy</a> though, so maybe a bit of an explanation of the process might help.</p>\n\n<p>If you're interested in where a particular protein is located, or how it changes location, you can use microscopy to investigate that. By using fluorescent dyes (or by creating fluorescent fusion proteins )that are attached specifically to the protein of interest (usually using a fluorescent-labelled antibody) you can visualise the location of the protein.</p>\n\n<p>By using different labels, that emit light of different wavelengths, you can tag multiple proteins in the cell at one time. These are then imaged in turn (using different filters), and the different images can be merged to show where the different proteins are in relation to each other, as in this image:</p>\n\n<p><img src=\"https://www.researchgate.net/profile/Lee_Eiden/publication/279308789/figure/fig1/AS:289266426626048@1445977898071/ATP-synthase-is-present-on-the-surface-of-mouse-pheochromocytoma-cells-Separate-confocal.png\" alt=\"Merged cell staining\">\nThe nucleus is often stained in blue as it is convenient to use a dye called <a href=\"https://www.thermofisher.com/uk/en/home/life-science/cell-analysis/fluorophores/dapi-stain.html\">DAPI</a>. The other references in these data are the endoplasmic reticulum - a 'factory' located next to the nucleus - in yellow, and microtubules - a component of the cell's 'skeleton' - in red.</p>\n\n<p>In this competition, it sounds as though you are to predict the subcellular location of the proteins tagged in green. The other three colours are provided to give you reference points of the cell structure. So, if the green protein is found in the same place (co-localises) with the blue, you might classify that protein as being in the nucleoplasm, or the nucleolus if it is in a smaller, more distinct structure within the nucleus.</p>\n\n<p>Hope this helps.</p>",
      "rawMarkdown": "I'm not taking part in this competition, so you might want to double check my answer in the context of the competition. I have done a good amount of cell staining and [fluorescence microscopy][1] though, so maybe a bit of an explanation of the process might help.\n\nIf you're interested in where a particular protein is located, or how it changes location, you can use microscopy to investigate that. By using fluorescent dyes (or by creating fluorescent fusion proteins )that are attached specifically to the protein of interest (usually using a fluorescent-labelled antibody) you can visualise the location of the protein.\n\nBy using different labels, that emit light of different wavelengths, you can tag multiple proteins in the cell at one time. These are then imaged in turn (using different filters), and the different images can be merged to show where the different proteins are in relation to each other, as in this image:\n\n![Merged cell staining][3]\n[3]: https://www.researchgate.net/profile/Lee_Eiden/publication/279308789/figure/fig1/AS:289266426626048@1445977898071/ATP-synthase-is-present-on-the-surface-of-mouse-pheochromocytoma-cells-Separate-confocal.png\n\nThe nucleus is often stained in blue as it is convenient to use a dye called [DAPI][2]. The other references in these data are the endoplasmic reticulum - a 'factory' located next to the nucleus - in yellow, and microtubules - a component of the cell's 'skeleton' - in red.\n\nIn this competition, it sounds as though you are to predict the subcellular location of the proteins tagged in green. The other three colours are provided to give you reference points of the cell structure. So, if the green protein is found in the same place (co-localises) with the blue, you might classify that protein as being in the nucleoplasm, or the nucleolus if it is in a smaller, more distinct structure within the nucleus.\n\nHope this helps.\n\n[1]: https://en.wikipedia.org/wiki/Fluorescence_microscope\n[2]: https://www.thermofisher.com/uk/en/home/life-science/cell-analysis/fluorophores/dapi-stain.html",
      "votes": null
    },
    {
      "id": "451711",
      "postDate": "01/07/2019 14:28:46",
      "content": "<p>WOW..THANKS SO MUCH! I konw what you mean. I totally understand what you are saying. </p>",
      "rawMarkdown": "WOW..THANKS SO MUCH! I konw what you mean. I totally understand what you are saying.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 450113,
      "author_name": "gomezp",
      "author_url": "",
      "post_date": "01/04/2019 10:28:29",
      "content": "<p>Afaik the different images stem from different stainings (<a href=\"https://en.wikipedia.org/wiki/Staining\">https://en.wikipedia.org/wiki/Staining</a> ) and therefore different things are visible in the images. All can give you indications on the correct label though.</p>",
      "votes": null,
      "replies": [
        {
          "id": 450564,
          "author_name": "vankexu",
          "author_url": "",
          "post_date": "01/05/2019 09:11:40",
          "content": "<p>Thanks, I know what you mean.Sorry for previous reply.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 450672,
      "author_name": "chrisbow",
      "author_url": "",
      "post_date": "01/05/2019 14:07:58",
      "content": "<p>I'm not taking part in this competition, so you might want to double check my answer in the context of the competition. I have done a good amount of cell staining and <a href=\"https://en.wikipedia.org/wiki/Fluorescence_microscope\">fluorescence microscopy</a> though, so maybe a bit of an explanation of the process might help.</p>\n\n<p>If you're interested in where a particular protein is located, or how it changes location, you can use microscopy to investigate that. By using fluorescent dyes (or by creating fluorescent fusion proteins )that are attached specifically to the protein of interest (usually using a fluorescent-labelled antibody) you can visualise the location of the protein.</p>\n\n<p>By using different labels, that emit light of different wavelengths, you can tag multiple proteins in the cell at one time. These are then imaged in turn (using different filters), and the different images can be merged to show where the different proteins are in relation to each other, as in this image:</p>\n\n<p><img src=\"https://www.researchgate.net/profile/Lee_Eiden/publication/279308789/figure/fig1/AS:289266426626048@1445977898071/ATP-synthase-is-present-on-the-surface-of-mouse-pheochromocytoma-cells-Separate-confocal.png\" alt=\"Merged cell staining\">\nThe nucleus is often stained in blue as it is convenient to use a dye called <a href=\"https://www.thermofisher.com/uk/en/home/life-science/cell-analysis/fluorophores/dapi-stain.html\">DAPI</a>. The other references in these data are the endoplasmic reticulum - a 'factory' located next to the nucleus - in yellow, and microtubules - a component of the cell's 'skeleton' - in red.</p>\n\n<p>In this competition, it sounds as though you are to predict the subcellular location of the proteins tagged in green. The other three colours are provided to give you reference points of the cell structure. So, if the green protein is found in the same place (co-localises) with the blue, you might classify that protein as being in the nucleoplasm, or the nucleolus if it is in a smaller, more distinct structure within the nucleus.</p>\n\n<p>Hope this helps.</p>",
      "votes": null,
      "replies": [
        {
          "id": 451711,
          "author_name": "vankexu",
          "author_url": "",
          "post_date": "01/07/2019 14:28:46",
          "content": "<p>WOW..THANKS SO MUCH! I konw what you mean. I totally understand what you are saying. </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "449100": "Here is the describation in '**What am I predicting**':\n\n All image samples are represented by four filters (stored as individual files), the protein of interest (green) plus three cellular landmarks: nucleus (blue), microtubules (red), endoplasmic reticulum (yellow). The green filter should hence be used to predict the label, and the other filters are used as references.\n\nI am confused about how to use these four pics correctly. What does 'The green filter should hence be used to predict the label, and the other filters are used as references'  mean? Thanks!! \n'",
    "450113": "Afaik the different images stem from different stainings (https://en.wikipedia.org/wiki/Staining ) and therefore different things are visible in the images. All can give you indications on the correct label though.",
    "450564": "Thanks, I know what you mean.Sorry for previous reply.",
    "450672": "I'm not taking part in this competition, so you might want to double check my answer in the context of the competition. I have done a good amount of cell staining and [fluorescence microscopy][1] though, so maybe a bit of an explanation of the process might help.\n\nIf you're interested in where a particular protein is located, or how it changes location, you can use microscopy to investigate that. By using fluorescent dyes (or by creating fluorescent fusion proteins )that are attached specifically to the protein of interest (usually using a fluorescent-labelled antibody) you can visualise the location of the protein.\n\nBy using different labels, that emit light of different wavelengths, you can tag multiple proteins in the cell at one time. These are then imaged in turn (using different filters), and the different images can be merged to show where the different proteins are in relation to each other, as in this image:\n\n![Merged cell staining][3]\n[3]: https://www.researchgate.net/profile/Lee_Eiden/publication/279308789/figure/fig1/AS:289266426626048@1445977898071/ATP-synthase-is-present-on-the-surface-of-mouse-pheochromocytoma-cells-Separate-confocal.png\n\nThe nucleus is often stained in blue as it is convenient to use a dye called [DAPI][2]. The other references in these data are the endoplasmic reticulum - a 'factory' located next to the nucleus - in yellow, and microtubules - a component of the cell's 'skeleton' - in red.\n\nIn this competition, it sounds as though you are to predict the subcellular location of the proteins tagged in green. The other three colours are provided to give you reference points of the cell structure. So, if the green protein is found in the same place (co-localises) with the blue, you might classify that protein as being in the nucleoplasm, or the nucleolus if it is in a smaller, more distinct structure within the nucleus.\n\nHope this helps.\n\n[1]: https://en.wikipedia.org/wiki/Fluorescence_microscope\n[2]: https://www.thermofisher.com/uk/en/home/life-science/cell-analysis/fluorophores/dapi-stain.html",
    "451711": "WOW..THANKS SO MUCH! I konw what you mean. I totally understand what you are saying."
  },
  "source": "meta"
}