https://www.kaggle.com/llttyy/calculate-the-gene-activity-score-of-multiome-data
\nHi all, here is an approach to transfer peaks information of scATAC-seq data into gene expression matrix, to reduce the noise exisiting in the original high-dim data. To run it you should install scanpy and episcanpy, and I recommend to use perosnal computer. That gene activity score matrix is different from the target we intend to predict because we did not consider some post-processing steps after transcirption. Hope it will be helpful for some people working on bio method. I haven't tried its performance but I will do it right now.
\nMoreover, it is very strange that I got 9200 genes in this step, but it seems that our target contains much more genes. Hope somebody can answer my question.
", "messages": [ { "id": "2012803", "postDate": "11/01/2022 12:38:26", "content": "https://www.kaggle.com/llttyy/calculate-the-gene-activity-score-of-multiome-data
\nHi all, here is an approach to transfer peaks information of scATAC-seq data into gene expression matrix, to reduce the noise exisiting in the original high-dim data. To run it you should install scanpy and episcanpy, and I recommend to use perosnal computer. That gene activity score matrix is different from the target we intend to predict because we did not consider some post-processing steps after transcirption. Hope it will be helpful for some people working on bio method. I haven't tried its performance but I will do it right now.
\nMoreover, it is very strange that I got 9200 genes in this step, but it seems that our target contains much more genes. Hope somebody can answer my question.
", "rawMarkdown": "https://www.kaggle.com/llttyy/calculate-the-gene-activity-score-of-multiome-data\n\nHi all, here is an approach to transfer peaks information of scATAC-seq data into gene expression matrix, to reduce the noise exisiting in the original high-dim data. To run it you should install scanpy and episcanpy, and I recommend to use perosnal computer. That gene activity score matrix is different from the target we intend to predict because we did not consider some post-processing steps after transcirption. Hope it will be helpful for some people working on bio method. I haven't tried its performance but I will do it right now. \n\nMoreover, it is very strange that I got 9200 genes in this step, but it seems that our target contains much more genes. Hope somebody can answer my question.", "votes": null }, { "id": "2013563", "postDate": "11/02/2022 02:51:40", "content": "Thanks for sharing. It's helpful @llttyy
\nKeep it up
Hmm we didnt try this yet ,thanks for sharing we will investigate hope it helps 👍..
", "rawMarkdown": "Hmm we didnt try this yet ,thanks for sharing we will investigate hope it helps 👍..", "votes": null }, { "id": "2014679", "postDate": "11/02/2022 18:35:00", "content": "Hi, thanks for your interests. I am still looking for the organizers to answer my questions in this part.
", "rawMarkdown": "Hi, thanks for your interests. I am still looking for the organizers to answer my questions in this part.", "votes": null }, { "id": "2015090", "postDate": "11/03/2022 04:56:05", "content": "You may want to look at this notebook and datasets for similar ideas and text on approach, with around 10,813 overlap genes with targets but also more than in targets.
\nhttps://www.kaggle.com/code/masato114/msci-multiome-using-geneactivity
\nhttps://www.kaggle.com/datasets/masato114/open-problems-train-geneactivity
\nhttps://www.kaggle.com/datasets/masato114/open-problems-test-geneactivity
Great, thanks a lot.
", "rawMarkdown": "Great, thanks a lot.", "votes": null }, { "id": "2015706", "postDate": "11/03/2022 13:39:00", "content": "Thanks for sharing! I learned more interesting stuff about Gene world 😄
\nPS. for the people who are looking for the GTF annotation files, I guess you can find through this link
@llttyy what's your question exactly?
\nThere are roughly 15,000 genes with expression values in the CITE GEX and roughly 20,000 in the Multiome GEX. However, it may be that not all these genes are being actively transcribed, so you may find only 9,200 genes with significant activity scores at their promoters.
\nI might infer that genes with accessibility at their promoters are going to be more actively transcribed (i.e. more RNA) at later time points.
", "rawMarkdown": "llttyy what's your question exactly?\n\nThere are roughly 15,000 genes with expression values in the CITE GEX and roughly 20,000 in the Multiome GEX. However, it may be that not all these genes are being actively transcribed, so you may find only 9,200 genes with significant activity scores at their promoters.\n\nI might infer that genes with accessibility at their promoters are going to be more actively transcribed (i.e. more RNA) at later time points.", "votes": null }, { "id": "2018554", "postDate": "11/05/2022 19:44:04", "content": "Ok… thanks for your sharing.
", "rawMarkdown": "Ok... thanks for your sharing.", "votes": null }, { "id": "2028958", "postDate": "11/14/2022 10:48:05", "content": "thank you so much, friend
", "rawMarkdown": "thank you so much, friend", "votes": null } ], "comments": [ { "id": 2013563, "author_name": "abhishek14398", "author_url": "", "post_date": "11/02/2022 02:51:40", "content": "Thanks for sharing. It's helpful @llttyy
\nKeep it up
Hmm we didnt try this yet ,thanks for sharing we will investigate hope it helps 👍..
", "votes": null, "replies": [ { "id": 2014679, "author_name": "llttyy", "author_url": "", "post_date": "11/02/2022 18:35:00", "content": "Hi, thanks for your interests. I am still looking for the organizers to answer my questions in this part.
", "votes": null, "replies": [] }, { "id": 2016144, "author_name": "danielburkhardt", "author_url": "", "post_date": "11/03/2022 20:08:47", "content": "@llttyy what's your question exactly?
\nThere are roughly 15,000 genes with expression values in the CITE GEX and roughly 20,000 in the Multiome GEX. However, it may be that not all these genes are being actively transcribed, so you may find only 9,200 genes with significant activity scores at their promoters.
\nI might infer that genes with accessibility at their promoters are going to be more actively transcribed (i.e. more RNA) at later time points.
", "votes": null, "replies": [] }, { "id": 2018554, "author_name": "llttyy", "author_url": "", "post_date": "11/05/2022 19:44:04", "content": "Ok… thanks for your sharing.
", "votes": null, "replies": [] } ] }, { "id": 2015090, "author_name": "something4kag", "author_url": "", "post_date": "11/03/2022 04:56:05", "content": "You may want to look at this notebook and datasets for similar ideas and text on approach, with around 10,813 overlap genes with targets but also more than in targets.
\nhttps://www.kaggle.com/code/masato114/msci-multiome-using-geneactivity
\nhttps://www.kaggle.com/datasets/masato114/open-problems-train-geneactivity
\nhttps://www.kaggle.com/datasets/masato114/open-problems-test-geneactivity
Great, thanks a lot.
", "votes": null, "replies": [] } ] }, { "id": 2015706, "author_name": "bwhale", "author_url": "", "post_date": "11/03/2022 13:39:00", "content": "Thanks for sharing! I learned more interesting stuff about Gene world 😄
\nPS. for the people who are looking for the GTF annotation files, I guess you can find through this link
thank you so much, friend
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2012803": "https://www.kaggle.com/llttyy/calculate-the-gene-activity-score-of-multiome-data\n\nHi all, here is an approach to transfer peaks information of scATAC-seq data into gene expression matrix, to reduce the noise exisiting in the original high-dim data. To run it you should install scanpy and episcanpy, and I recommend to use perosnal computer. That gene activity score matrix is different from the target we intend to predict because we did not consider some post-processing steps after transcirption. Hope it will be helpful for some people working on bio method. I haven't tried its performance but I will do it right now. \n\nMoreover, it is very strange that I got 9200 genes in this step, but it seems that our target contains much more genes. Hope somebody can answer my question.", "2013563": "Thanks for sharing. It's helpful @llttyy \nKeep it up", "2014579": "Hmm we didnt try this yet ,thanks for sharing we will investigate hope it helps 👍..", "2014679": "Hi, thanks for your interests. I am still looking for the organizers to answer my questions in this part.", "2015090": "You may want to look at this notebook and datasets for similar ideas and text on approach, with around 10,813 overlap genes with targets but also more than in targets.\n\nhttps://www.kaggle.com/code/masato114/msci-multiome-using-geneactivity\n\nhttps://www.kaggle.com/datasets/masato114/open-problems-train-geneactivity\nhttps://www.kaggle.com/datasets/masato114/open-problems-test-geneactivity", "2015585": "Great, thanks a lot.", "2015706": "Thanks for sharing! I learned more interesting stuff about Gene world 😄\nPS. for the people who are looking for the GTF annotation files, I guess you can find through [this link](https://www.gencodegenes.org/human/)", "2016144": "llttyy what's your question exactly?\n\nThere are roughly 15,000 genes with expression values in the CITE GEX and roughly 20,000 in the Multiome GEX. However, it may be that not all these genes are being actively transcribed, so you may find only 9,200 genes with significant activity scores at their promoters.\n\nI might infer that genes with accessibility at their promoters are going to be more actively transcribed (i.e. more RNA) at later time points.", "2018554": "Ok... thanks for your sharing.", "2028958": "thank you so much, friend" }, "source": "meta" }