{ "id": 361814, "title": "Open Problems DNA/Single Cells. Join this Unique Bioinformatics Competition.", "url": "/competitions/open-problems-multimodal/discussion/361814", "author_name": "", "post_date": "2022-10-23T22:25:42.442608700Z", "votes": 15, "comment_count": 7, "views": 0, "content": "

We have still 23 days more. Clock is ticking. Deadline is near.

\n

According to A. Chervov (Bioinformatician at Institute Curie - Paris) : \"The data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\"

\n

D. Burkhardt (Competition Host): \"This competition is about predicting the relationship between different genetic measurements (DNA, RNA, and protein) measured simultaneously in single cells.\"

\n

Therefore, don't be shy if you don't have experience in that subject. In this competition HDF5 (h5) file is being used too. I've already made an effort and tried one Kaggle Notebook to learn a little bit more.

\n

Attention Newcomers

\n

Adding some notes that will be helpful for you Newcomers:

\n

\"For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\"

\n

There are two competitions in one: Cite and Multiome

\n

https://www.kaggle.com/competitions/open-problems-multimodal/discussion/360180
\nBy Elias (gehallak 24th)

\n

\"Both are multi target (the targets have multiple columns)
\nBoth have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.\"

\n

Metric: TARGETWISE

\n

\"Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). See discussion on possible tricks.\"

\n

\"Private and Public LB are different. That might cause a shake-up: problem, possible solution - validation scheme. Some other ideas, problems collected: 1, 2,3.\"
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253
\nBy A. Chervov (alexandervc)

\n

\"Bioinformatics is one such interdisciplinary field that stands with of Bio-Sciences, Information Technology, Mathematics, and Statistics. This overlapping area provides a tremendous advantage over others options and that's why it became a golden opportunity for current and future generations.\"

\n

https://uem.edu.in/uem-kolkata-blog/2021/06/23/future-career-scope-of-bioinformatics/

\n

Join it. Make some public code. Embrace this promising field of Data Science.

", "messages": [ { "id": "2001287", "postDate": "10/23/2022 22:25:42", "content": "

We have still 23 days more. Clock is ticking. Deadline is near.

\n

According to A. Chervov (Bioinformatician at Institute Curie - Paris) : \"The data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\"

\n

D. Burkhardt (Competition Host): \"This competition is about predicting the relationship between different genetic measurements (DNA, RNA, and protein) measured simultaneously in single cells.\"

\n

Therefore, don't be shy if you don't have experience in that subject. In this competition HDF5 (h5) file is being used too. I've already made an effort and tried one Kaggle Notebook to learn a little bit more.

\n

Attention Newcomers

\n

Adding some notes that will be helpful for you Newcomers:

\n

\"For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\"

\n

There are two competitions in one: Cite and Multiome

\n

https://www.kaggle.com/competitions/open-problems-multimodal/discussion/360180
\nBy Elias (gehallak 24th)

\n

\"Both are multi target (the targets have multiple columns)
\nBoth have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.\"

\n

Metric: TARGETWISE

\n

\"Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). See discussion on possible tricks.\"

\n

\"Private and Public LB are different. That might cause a shake-up: problem, possible solution - validation scheme. Some other ideas, problems collected: 1, 2,3.\"
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253
\nBy A. Chervov (alexandervc)

\n

\"Bioinformatics is one such interdisciplinary field that stands with of Bio-Sciences, Information Technology, Mathematics, and Statistics. This overlapping area provides a tremendous advantage over others options and that's why it became a golden opportunity for current and future generations.\"

\n

https://uem.edu.in/uem-kolkata-blog/2021/06/23/future-career-scope-of-bioinformatics/

\n

Join it. Make some public code. Embrace this promising field of Data Science.

", "rawMarkdown": "We have still 23 days more. Clock is ticking. Deadline is near.\n\nAccording to A. Chervov (Bioinformatician at Institute Curie - Paris) : \"The data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\"\n\nD. Burkhardt (Competition Host): \"This competition is about predicting the relationship between different genetic measurements (DNA, RNA, and protein) measured simultaneously in single cells.\"\n\nTherefore, don't be shy if you don't have experience in that subject. In this competition HDF5 (h5) file is being used too. I've already made an effort and tried one Kaggle Notebook to learn a little bit more.\n\n#Attention Newcomers\n\nAdding some notes that will be helpful for you Newcomers:\n\n\"For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\"\n\n#There are two competitions in one: Cite and Multiome\n\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360180\nBy Elias (gehallak 24th)\n\n\"Both are multi target (the targets have multiple columns)\nBoth have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.\"\n\n#Metric: TARGETWISE\n\n \"Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). See discussion on possible tricks.\"\n\n\"Private and Public LB are different. That might cause a shake-up: problem, possible solution - validation scheme. Some other ideas, problems collected: 1, 2,3.\"\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253\nBy A. Chervov (alexandervc)\n\n\"Bioinformatics is one such interdisciplinary field that stands with of Bio-Sciences, Information Technology, Mathematics, and Statistics. This overlapping area provides a tremendous advantage over others options and that's why it became a golden opportunity for current and future generations.\"\n\nhttps://uem.edu.in/uem-kolkata-blog/2021/06/23/future-career-scope-of-bioinformatics/\n\n#Join it. Make some public code. Embrace this promising field of Data Science.", "votes": null }, { "id": "2001490", "postDate": "10/24/2022 04:57:06", "content": "

For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.

\n", "rawMarkdown": "For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\n- There are two competitions in one: Cite and Multiome\n- Both are multi target (the targets have multiple columns)\n- Both have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.", "votes": null }, { "id": "2001627", "postDate": "10/24/2022 07:29:03", "content": "

Following Elias, let me add couple of notes, which might be useful for newcomers, to pay attention on:

\n\n

Welcome to join !

", "rawMarkdown": "Following Elias, let me add couple of notes, which might be useful for newcomers, to pay attention on: \n\n- Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). [See discussion on possible tricks](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253). \n- Private and Public LB are different. That might cause a shake-up: [problem](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/347202), possible solution -[ validation scheme](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/358860)\n- Some other ideas, problems collected: [1](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025), [2](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311),[3](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/353225).\n\nWelcome to join !", "votes": null }, { "id": "2001942", "postDate": "10/24/2022 12:19:06", "content": "

Thank you Elias. I added your worthy comment on the topic.
\nThank you also for your insightful contributions on this Open Modal single cell competition.

", "rawMarkdown": "Thank you Elias. I added your worthy comment on the topic. \nThank you also for your insightful contributions on this Open Modal single cell competition.", "votes": null }, { "id": "2001956", "postDate": "10/24/2022 12:27:38", "content": "

Hi Alexander, I added your comment with the link of your topic about the metric.

\n

As I wrote above, you and Elias are giving a remarkable contribution to this Single cell/DNA measurements competition.

", "rawMarkdown": "Hi Alexander, I added your comment with the link of your topic about the metric.\n\nAs I wrote above, you and Elias are giving a remarkable contribution to this Single cell/DNA measurements competition.", "votes": null }, { "id": "2002115", "postDate": "10/24/2022 14:19:08", "content": "

2 in 1 bioinformatics competition. Challenging but very interesting

", "rawMarkdown": "2 in 1 bioinformatics competition. Challenging but very interesting", "votes": null }, { "id": "2002118", "postDate": "10/24/2022 14:21:11", "content": "

Looking for a merger to form a winning team

", "rawMarkdown": "Looking for a merger to form a winning team", "votes": null }, { "id": "2002183", "postDate": "10/24/2022 15:15:01", "content": "

Hi Thierrynakeng

\n

Due to my lack of knowledge I'm not able to merge in any team. Imagine a winning team : )

\n

I hope you can find anyone to merge with in that challenge. Or even share your insights about the competition.

", "rawMarkdown": "Hi Thierrynakeng \n\nDue to my lack of knowledge I'm not able to merge in any team. Imagine a winning team : )\n\nI hope you can find anyone to merge with in that challenge. Or even share your insights about the competition.", "votes": null } ], "comments": [ { "id": 2001490, "author_name": "gehallak", "author_url": "", "post_date": "10/24/2022 04:57:06", "content": "

For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.

\n", "votes": null, "replies": [ { "id": 2001942, "author_name": "mpwolke", "author_url": "", "post_date": "10/24/2022 12:19:06", "content": "

Thank you Elias. I added your worthy comment on the topic.
\nThank you also for your insightful contributions on this Open Modal single cell competition.

", "votes": null, "replies": [] } ] }, { "id": 2001627, "author_name": "alexandervc", "author_url": "", "post_date": "10/24/2022 07:29:03", "content": "

Following Elias, let me add couple of notes, which might be useful for newcomers, to pay attention on:

\n\n

Welcome to join !

", "votes": null, "replies": [ { "id": 2001956, "author_name": "mpwolke", "author_url": "", "post_date": "10/24/2022 12:27:38", "content": "

Hi Alexander, I added your comment with the link of your topic about the metric.

\n

As I wrote above, you and Elias are giving a remarkable contribution to this Single cell/DNA measurements competition.

", "votes": null, "replies": [] } ] }, { "id": 2002115, "author_name": "thierrynankeng", "author_url": "", "post_date": "10/24/2022 14:19:08", "content": "

2 in 1 bioinformatics competition. Challenging but very interesting

", "votes": null, "replies": [] }, { "id": 2002118, "author_name": "thierrynankeng", "author_url": "", "post_date": "10/24/2022 14:21:11", "content": "

Looking for a merger to form a winning team

", "votes": null, "replies": [ { "id": 2002183, "author_name": "mpwolke", "author_url": "", "post_date": "10/24/2022 15:15:01", "content": "

Hi Thierrynakeng

\n

Due to my lack of knowledge I'm not able to merge in any team. Imagine a winning team : )

\n

I hope you can find anyone to merge with in that challenge. Or even share your insights about the competition.

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2001287": "We have still 23 days more. Clock is ticking. Deadline is near.\n\nAccording to A. Chervov (Bioinformatician at Institute Curie - Paris) : \"The data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\"\n\nD. Burkhardt (Competition Host): \"This competition is about predicting the relationship between different genetic measurements (DNA, RNA, and protein) measured simultaneously in single cells.\"\n\nTherefore, don't be shy if you don't have experience in that subject. In this competition HDF5 (h5) file is being used too. I've already made an effort and tried one Kaggle Notebook to learn a little bit more.\n\n#Attention Newcomers\n\nAdding some notes that will be helpful for you Newcomers:\n\n\"For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\"\n\n#There are two competitions in one: Cite and Multiome\n\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360180\nBy Elias (gehallak 24th)\n\n\"Both are multi target (the targets have multiple columns)\nBoth have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.\"\n\n#Metric: TARGETWISE\n\n \"Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). See discussion on possible tricks.\"\n\n\"Private and Public LB are different. That might cause a shake-up: problem, possible solution - validation scheme. Some other ideas, problems collected: 1, 2,3.\"\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253\nBy A. Chervov (alexandervc)\n\n\"Bioinformatics is one such interdisciplinary field that stands with of Bio-Sciences, Information Technology, Mathematics, and Statistics. This overlapping area provides a tremendous advantage over others options and that's why it became a golden opportunity for current and future generations.\"\n\nhttps://uem.edu.in/uem-kolkata-blog/2021/06/23/future-career-scope-of-bioinformatics/\n\n#Join it. Make some public code. Embrace this promising field of Data Science.", "2001490": "For those of you who worked on the Amex competition, it is somehow similar (tabular data), but there are some major differences.\n- There are two competitions in one: Cite and Multiome\n- Both are multi target (the targets have multiple columns)\n- Both have a very large number of features. In Multiome, the number of features is even larger than the number of samples (more columns than rows in train and test). This is an additional difficulty which probably often occurs in bioinformatics.", "2001627": "Following Elias, let me add couple of notes, which might be useful for newcomers, to pay attention on: \n\n- Metric is - TARGETWISE correlation with afterwards average by samples. Not usual correlation along the samples (!). [See discussion on possible tricks](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/360253). \n- Private and Public LB are different. That might cause a shake-up: [problem](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/347202), possible solution -[ validation scheme](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/358860)\n- Some other ideas, problems collected: [1](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025), [2](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311),[3](https://www.kaggle.com/competitions/open-problems-multimodal/discussion/353225).\n\nWelcome to join !", "2001942": "Thank you Elias. I added your worthy comment on the topic. \nThank you also for your insightful contributions on this Open Modal single cell competition.", "2001956": "Hi Alexander, I added your comment with the link of your topic about the metric.\n\nAs I wrote above, you and Elias are giving a remarkable contribution to this Single cell/DNA measurements competition.", "2002115": "2 in 1 bioinformatics competition. Challenging but very interesting", "2002118": "Looking for a merger to form a winning team", "2002183": "Hi Thierrynakeng \n\nDue to my lack of knowledge I'm not able to merge in any team. Imagine a winning team : )\n\nI hope you can find anyone to merge with in that challenge. Or even share your insights about the competition." }, "source": "meta" }