{
  "id": 453376,
  "title": "Missing data for DMSO in adata",
  "url": "/competitions/open-problems-single-cell-perturbations/discussion/453376",
  "author_name": "",
  "post_date": "2023-11-06T06:12:01.299975500Z",
  "votes": 13,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Dear competition hosts, could you please explain the absence of data for DMSO for several genes in some cell types? I have just tried to aggregate DMSO counts from the adata_train file and found that there are missing DMSO effects on 2485 genes (at least one in each cell type), while there are data for other drugs on these genes. </p>\n<p>For example, if you subset only cells treated with DMSO you can see that there is no data for COL4A2-AS2 gene - completely missing.</p>\n<p>My guess is as you told previously on missing data in the multiome dataset, missing effects here means that there were 0 counts. Is it correct?</p>\n<p>Also, at what step of data preparation you did subtract the effects of DMSO from the drug effect? And how did you do this with missing data for DMSO, if they are not zeros?</p>",
  "messages": [
    {
      "id": "2514218",
      "postDate": "11/06/2023 06:12:01",
      "content": "<p>Dear competition hosts, could you please explain the absence of data for DMSO for several genes in some cell types? I have just tried to aggregate DMSO counts from the adata_train file and found that there are missing DMSO effects on 2485 genes (at least one in each cell type), while there are data for other drugs on these genes. </p>\n<p>For example, if you subset only cells treated with DMSO you can see that there is no data for COL4A2-AS2 gene - completely missing.</p>\n<p>My guess is as you told previously on missing data in the multiome dataset, missing effects here means that there were 0 counts. Is it correct?</p>\n<p>Also, at what step of data preparation you did subtract the effects of DMSO from the drug effect? And how did you do this with missing data for DMSO, if they are not zeros?</p>",
      "rawMarkdown": "Dear competition hosts, could you please explain the absence of data for DMSO for several genes in some cell types? I have just tried to aggregate DMSO counts from the adata_train file and found that there are missing DMSO effects on 2485 genes (at least one in each cell type), while there are data for other drugs on these genes. \n\nFor example, if you subset only cells treated with DMSO you can see that there is no data for COL4A2-AS2 gene - completely missing.\n\nMy guess is as you told previously on missing data in the multiome dataset, missing effects here means that there were 0 counts. Is it correct?\n\nAlso, at what step of data preparation you did subtract the effects of DMSO from the drug effect? And how did you do this with missing data for DMSO, if they are not zeros?",
      "votes": null
    },
    {
      "id": "2517607",
      "postDate": "11/08/2023 16:13:34",
      "content": "<p><a href=\"https://www.kaggle.com/danielburkhardt\" target=\"_blank\">@danielburkhardt</a> , <a href=\"https://www.kaggle.com/andrewbenz\" target=\"_blank\">@andrewbenz</a>  may be you can comment on this question, please? </p>",
      "rawMarkdown": "danielburkhardt , @andrewbenz  may be you can comment on this question, please?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2517607,
      "author_name": "antoninadolgorukova",
      "author_url": "",
      "post_date": "11/08/2023 16:13:34",
      "content": "<p><a href=\"https://www.kaggle.com/danielburkhardt\" target=\"_blank\">@danielburkhardt</a> , <a href=\"https://www.kaggle.com/andrewbenz\" target=\"_blank\">@andrewbenz</a>  may be you can comment on this question, please? </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2514218": "Dear competition hosts, could you please explain the absence of data for DMSO for several genes in some cell types? I have just tried to aggregate DMSO counts from the adata_train file and found that there are missing DMSO effects on 2485 genes (at least one in each cell type), while there are data for other drugs on these genes. \n\nFor example, if you subset only cells treated with DMSO you can see that there is no data for COL4A2-AS2 gene - completely missing.\n\nMy guess is as you told previously on missing data in the multiome dataset, missing effects here means that there were 0 counts. Is it correct?\n\nAlso, at what step of data preparation you did subtract the effects of DMSO from the drug effect? And how did you do this with missing data for DMSO, if they are not zeros?",
    "2517607": "danielburkhardt , @andrewbenz  may be you can comment on this question, please?"
  },
  "source": "meta"
}