{ "id": 460378, "title": "Thanks from the hosts, and next steps!", "url": "/competitions/stanford-ribonanza-rna-folding/discussion/460378", "author_name": "", "post_date": "2023-12-09T01:15:05.269075700Z", "votes": 20, "comment_count": 9, "views": 0, "content": "

Congratulations to all who made submissions to the Stanford Ribonanza RNA Folding challenge!

\n

It has been exciting for us hosts to engage with all 900 of you through forum posts, features, and through your models.

\n

Many of our questions ahead of the competition about architectures, pseudolabeling, external data, and more appear to have been answered.

\n

And we are excited to see what happens as your models become publicly released for use by the broader RNA science community.

\n

There will be more discussion over the next few days and follow-on seminars and preprints. This will start with a summary talk on Ribonanza 2 next week on Friday Dec. 15, at the MLSB workshop at NeurIPS 2023 — hope to see some of you there, and we’ll make video available after that conference.

\n

In addition, for the many of you who have become hooked on RNA, we hope you’ll continue to advance our understanding of these important molecules through three forums:

\n
    \n

  1. Kaggle. Please do post a summary of your model and lessons learned here in discussion, with links to public Kaggle notebooks. These resources will allow us, the broader science community, and future Kagglers, to best build on your collective advances.

    2. \n

  2. CASP16. A different and very high profile form of evaluation of RNA structure models involves comparing predicted 3D coordinates of selected RNA molecules to to experimentally determined structures. This task is evaluated very two years in the CASP competitions. The next one, CASP16, will start in April 2024 and, like the last CASP15, there will be an RNA category. Hope to see some of you there!

    3. \n

  3. Ribonanza2. We are looking to further scale up our wet-lab experiments to probe 10x more sequences — or perhaps even more. We’re tentatively calling this initiative the “Ribonanza2” project. We’d love sequences drawn from the most diverse sources possible, and we invite Kaggle teams to contribute. If there are sequences that you think would best teach your models how to get better, we would like to make them and map them! Teams who sign up at this form will have until January 15, 2024 to submit up to 100,000 sequences.

    4. \n
\n

Thanks to all who participated - and see you in the RNA world!

\n

Rhiju Das, for the hosts

\n

Hosts
\nRhiju Das @rhijudas
\nShujun He @shujun717
\nThomas Karagianes @brainbowrna
\nJill Townley @digitalembrace
\nRachael Kretsch @rkretsch
\nGrace Nye @gracenye8
\nRui Huang @RuiHuang24

\n

Thanks to members of the Eterna community and the Das laboratory, including John Nicol, Christian Choe @HCL-Rantig, and Jonathan Romano @jonathanromano, for key contributions in software development and library design.

\n

And thanks to our collaborators at Kaggle, with special acknowledgments to:

\n

Maggie Demkin @maggiemd
\nInversion @inversion
\nAshley Chow @ashleychow

", "messages": [ { "id": "2554254", "postDate": "12/09/2023 01:15:05", "content": "

Congratulations to all who made submissions to the Stanford Ribonanza RNA Folding challenge!

\n

It has been exciting for us hosts to engage with all 900 of you through forum posts, features, and through your models.

\n

Many of our questions ahead of the competition about architectures, pseudolabeling, external data, and more appear to have been answered.

\n

And we are excited to see what happens as your models become publicly released for use by the broader RNA science community.

\n

There will be more discussion over the next few days and follow-on seminars and preprints. This will start with a summary talk on Ribonanza 2 next week on Friday Dec. 15, at the MLSB workshop at NeurIPS 2023 — hope to see some of you there, and we’ll make video available after that conference.

\n

In addition, for the many of you who have become hooked on RNA, we hope you’ll continue to advance our understanding of these important molecules through three forums:

\n
    \n

  1. Kaggle. Please do post a summary of your model and lessons learned here in discussion, with links to public Kaggle notebooks. These resources will allow us, the broader science community, and future Kagglers, to best build on your collective advances.

    2. \n

  2. CASP16. A different and very high profile form of evaluation of RNA structure models involves comparing predicted 3D coordinates of selected RNA molecules to to experimentally determined structures. This task is evaluated very two years in the CASP competitions. The next one, CASP16, will start in April 2024 and, like the last CASP15, there will be an RNA category. Hope to see some of you there!

    3. \n

  3. Ribonanza2. We are looking to further scale up our wet-lab experiments to probe 10x more sequences — or perhaps even more. We’re tentatively calling this initiative the “Ribonanza2” project. We’d love sequences drawn from the most diverse sources possible, and we invite Kaggle teams to contribute. If there are sequences that you think would best teach your models how to get better, we would like to make them and map them! Teams who sign up at this form will have until January 15, 2024 to submit up to 100,000 sequences.

    4. \n
\n

Thanks to all who participated - and see you in the RNA world!

\n

Rhiju Das, for the hosts

\n

Hosts
\nRhiju Das @rhijudas
\nShujun He @shujun717
\nThomas Karagianes @brainbowrna
\nJill Townley @digitalembrace
\nRachael Kretsch @rkretsch
\nGrace Nye @gracenye8
\nRui Huang @RuiHuang24

\n

Thanks to members of the Eterna community and the Das laboratory, including John Nicol, Christian Choe @HCL-Rantig, and Jonathan Romano @jonathanromano, for key contributions in software development and library design.

\n

And thanks to our collaborators at Kaggle, with special acknowledgments to:

\n

Maggie Demkin @maggiemd
\nInversion @inversion
\nAshley Chow @ashleychow

", "rawMarkdown": "Congratulations to all who made submissions to the Stanford Ribonanza RNA Folding challenge!\n\nIt has been exciting for us hosts to engage with all 900 of you through forum posts, features, and through your models.\n\nMany of our questions ahead of the competition about architectures, pseudolabeling, external data, and more appear to have been answered.\n\nAnd we are excited to see what happens as your models become publicly released for use by the broader RNA science community.\n\nThere will be more discussion over the next few days and follow-on seminars and preprints. This will start with a summary talk on Ribonanza 2 next week on Friday Dec. 15, at the MLSB workshop at NeurIPS 2023 — hope to see some of you there, and we’ll make video available after that conference.\n\nIn addition, for the many of you who have become hooked on RNA, we hope you’ll continue to advance our understanding of these important molecules through three forums:\n\n1. **Kaggle**. Please do post a summary of your model and lessons learned here in discussion, with links to public Kaggle notebooks. These resources will allow us, the broader science community, and future Kagglers, to best build on your collective advances.\n\n2. **CASP16**. A different and very high profile form of evaluation of RNA structure models involves comparing predicted 3D coordinates of selected RNA molecules to to experimentally determined structures. This task is evaluated very two years in the [CASP competitions](https://predictioncenter.org/). The next one, CASP16, will start in April 2024 and, like the last CASP15, there will be an RNA category. Hope to see some of you there!\n\n3. **Ribonanza2**. We are looking to further scale up our wet-lab experiments to probe 10x more sequences — or perhaps even more. We’re tentatively calling this initiative the “Ribonanza2” project. We’d love sequences drawn from the most diverse sources possible, and we invite Kaggle teams to contribute. If there are sequences that you think would best teach your models how to get better, we would like to make them and map them! Teams who sign up at [this form](https://forms.gle/sfBWnmxXHuaTHdE58) will have until January 15, 2024 to submit up to 100,000 sequences.\n\nThanks to all who participated - and see you in the RNA world!\n\nRhiju Das, for the hosts\n\n**Hosts**\nRhiju Das @rhijudas\nShujun He @shujun717\nThomas Karagianes @brainbowrna\nJill Townley @digitalembrace\nRachael Kretsch @rkretsch\nGrace Nye @gracenye8 \nRui Huang @RuiHuang24\n\nThanks to members of the Eterna community and the Das laboratory, including John Nicol, Christian Choe @HCL-Rantig, and Jonathan Romano @jonathanromano, for key contributions in software development and library design.\n\nAnd thanks to our collaborators at Kaggle, with special acknowledgments to:\n\nMaggie Demkin @maggiemd\nInversion @inversion\nAshley Chow @ashleychow", "votes": null }, { "id": "2554340", "postDate": "12/09/2023 04:26:56", "content": "

Preprints - are you going to publish a paper/series of papers on the results of this competition? I failed to find any information about that, thanks!

\n

3 - could you specify the size of sequences we can submit?

", "rawMarkdown": "**Preprints** - are you going to publish a paper/series of papers on the results of this competition? I failed to find any information about that, thanks!\n\n\n3 - could you specify the size of sequences we can submit?", "votes": null }, { "id": "2554342", "postDate": "12/09/2023 04:31:35", "content": "

Thank you for hosting the competition!

\n

Is it permissible to post the link to the code in a public discussion here on Kaggle, however, the link would lead to a GitHub repository rather than a Kaggle notebook? I would like to share my code, but it is in a repository and is organized as multiple python scripts. (Currently, the repository is private.)

", "rawMarkdown": "Thank you for hosting the competition!\n\nIs it permissible to post the link to the code in a public discussion here on Kaggle, however, the link would lead to a GitHub repository rather than a Kaggle notebook? I would like to share my code, but it is in a repository and is organized as multiple python scripts. (Currently, the repository is private.)", "votes": null }, { "id": "2554982", "postDate": "12/09/2023 15:33:23", "content": "

Yes, a paper will be written on the competition results.

\n

3 - The length of the sequences is 100 nucleotides. The leader, barcode, and tail will be added by the Das Lab.

", "rawMarkdown": "Yes, a paper will be written on the competition results.\n\n3 - The length of the sequences is 100 nucleotides. The leader, barcode, and tail will be added by the Das Lab.", "votes": null }, { "id": "2555127", "postDate": "12/09/2023 17:49:23", "content": "

Are you going to include authors of best solutions as authors of the paper?

\n

Our team will be happy to contribute to the paper

", "rawMarkdown": "Are you going to include authors of best solutions as authors of the paper? \n\nOur team will be happy to contribute to the paper", "votes": null }, { "id": "2555260", "postDate": "12/09/2023 19:14:42", "content": "

Yes, definitely OK to post on GitHub!

", "rawMarkdown": "Yes, definitely OK to post on GitHub!", "votes": null }, { "id": "2555263", "postDate": "12/09/2023 19:21:18", "content": "

That's great to hear!

\n

Also please note that there is no restriction on your team or other teams writing additional publications, if you have the bandwidth.

\n

Looking forward to more discussions as we put together publications.

", "rawMarkdown": "That's great to hear! \n\nAlso please note that there is no restriction on your team or other teams writing additional publications, if you have the bandwidth. \n\nLooking forward to more discussions as we put together publications.", "votes": null }, { "id": "2555265", "postDate": "12/09/2023 19:27:46", "content": "

Thank you!

", "rawMarkdown": "Thank you!", "votes": null }, { "id": "2555270", "postDate": "12/09/2023 19:36:48", "content": "

@rhijudas was that a 'yes'? 😄

\n

Anyway, +1 to @dmitrypenzar1996, I'd also appreciate if @rhijudas / @digitalembrace (or one of other organizers not present in this thread) could clarify whether the best solutions' authors would be able to co-author the flagship paper. Really excited to participate!

", "rawMarkdown": "rhijudas was that a 'yes'? 😄\n\nAnyway, +1 to @dmitrypenzar1996, I'd also appreciate if @rhijudas / @digitalembrace (or one of other organizers not present in this thread) could clarify whether the best solutions' authors would be able to co-author the flagship paper. Really excited to participate!", "votes": null }, { "id": "2560446", "postDate": "12/13/2023 16:36:33", "content": "

@digitalembrace could you clarify this step? This also influence what we are going to investigate in our additional publication

", "rawMarkdown": "digitalembrace could you clarify this step? This also influence what we are going to investigate in our additional publication", "votes": null } ], "comments": [ { "id": 2554340, "author_name": "dmitrypenzar1996", "author_url": "", "post_date": "12/09/2023 04:26:56", "content": "

Preprints - are you going to publish a paper/series of papers on the results of this competition? I failed to find any information about that, thanks!

\n

3 - could you specify the size of sequences we can submit?

", "votes": null, "replies": [ { "id": 2554982, "author_name": "digitalembrace", "author_url": "", "post_date": "12/09/2023 15:33:23", "content": "

Yes, a paper will be written on the competition results.

\n

3 - The length of the sequences is 100 nucleotides. The leader, barcode, and tail will be added by the Das Lab.

", "votes": null, "replies": [ { "id": 2555127, "author_name": "dmitrypenzar1996", "author_url": "", "post_date": "12/09/2023 17:49:23", "content": "

Are you going to include authors of best solutions as authors of the paper?

\n

Our team will be happy to contribute to the paper

", "votes": null, "replies": [ { "id": 2555263, "author_name": "rhijudas", "author_url": "", "post_date": "12/09/2023 19:21:18", "content": "

That's great to hear!

\n

Also please note that there is no restriction on your team or other teams writing additional publications, if you have the bandwidth.

\n

Looking forward to more discussions as we put together publications.

", "votes": null, "replies": [ { "id": 2555270, "author_name": "polymerase", "author_url": "", "post_date": "12/09/2023 19:36:48", "content": "

@rhijudas was that a 'yes'? 😄

\n

Anyway, +1 to @dmitrypenzar1996, I'd also appreciate if @rhijudas / @digitalembrace (or one of other organizers not present in this thread) could clarify whether the best solutions' authors would be able to co-author the flagship paper. Really excited to participate!

", "votes": null, "replies": [ { "id": 2560446, "author_name": "dmitrypenzar1996", "author_url": "", "post_date": "12/13/2023 16:36:33", "content": "

@digitalembrace could you clarify this step? This also influence what we are going to investigate in our additional publication

", "votes": null, "replies": [] } ] } ] } ] } ] } ] }, { "id": 2554342, "author_name": "ilyanaanderson", "author_url": "", "post_date": "12/09/2023 04:31:35", "content": "

Thank you for hosting the competition!

\n

Is it permissible to post the link to the code in a public discussion here on Kaggle, however, the link would lead to a GitHub repository rather than a Kaggle notebook? I would like to share my code, but it is in a repository and is organized as multiple python scripts. (Currently, the repository is private.)

", "votes": null, "replies": [ { "id": 2555260, "author_name": "rhijudas", "author_url": "", "post_date": "12/09/2023 19:14:42", "content": "

Yes, definitely OK to post on GitHub!

", "votes": null, "replies": [ { "id": 2555265, "author_name": "ilyanaanderson", "author_url": "", "post_date": "12/09/2023 19:27:46", "content": "

Thank you!

", "votes": null, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "2554254": "Congratulations to all who made submissions to the Stanford Ribonanza RNA Folding challenge!\n\nIt has been exciting for us hosts to engage with all 900 of you through forum posts, features, and through your models.\n\nMany of our questions ahead of the competition about architectures, pseudolabeling, external data, and more appear to have been answered.\n\nAnd we are excited to see what happens as your models become publicly released for use by the broader RNA science community.\n\nThere will be more discussion over the next few days and follow-on seminars and preprints. This will start with a summary talk on Ribonanza 2 next week on Friday Dec. 15, at the MLSB workshop at NeurIPS 2023 — hope to see some of you there, and we’ll make video available after that conference.\n\nIn addition, for the many of you who have become hooked on RNA, we hope you’ll continue to advance our understanding of these important molecules through three forums:\n\n1. **Kaggle**. Please do post a summary of your model and lessons learned here in discussion, with links to public Kaggle notebooks. These resources will allow us, the broader science community, and future Kagglers, to best build on your collective advances.\n\n2. **CASP16**. A different and very high profile form of evaluation of RNA structure models involves comparing predicted 3D coordinates of selected RNA molecules to to experimentally determined structures. This task is evaluated very two years in the [CASP competitions](https://predictioncenter.org/). The next one, CASP16, will start in April 2024 and, like the last CASP15, there will be an RNA category. Hope to see some of you there!\n\n3. **Ribonanza2**. We are looking to further scale up our wet-lab experiments to probe 10x more sequences — or perhaps even more. We’re tentatively calling this initiative the “Ribonanza2” project. We’d love sequences drawn from the most diverse sources possible, and we invite Kaggle teams to contribute. If there are sequences that you think would best teach your models how to get better, we would like to make them and map them! Teams who sign up at [this form](https://forms.gle/sfBWnmxXHuaTHdE58) will have until January 15, 2024 to submit up to 100,000 sequences.\n\nThanks to all who participated - and see you in the RNA world!\n\nRhiju Das, for the hosts\n\n**Hosts**\nRhiju Das @rhijudas\nShujun He @shujun717\nThomas Karagianes @brainbowrna\nJill Townley @digitalembrace\nRachael Kretsch @rkretsch\nGrace Nye @gracenye8 \nRui Huang @RuiHuang24\n\nThanks to members of the Eterna community and the Das laboratory, including John Nicol, Christian Choe @HCL-Rantig, and Jonathan Romano @jonathanromano, for key contributions in software development and library design.\n\nAnd thanks to our collaborators at Kaggle, with special acknowledgments to:\n\nMaggie Demkin @maggiemd\nInversion @inversion\nAshley Chow @ashleychow", "2554340": "**Preprints** - are you going to publish a paper/series of papers on the results of this competition? I failed to find any information about that, thanks!\n\n\n3 - could you specify the size of sequences we can submit?", "2554342": "Thank you for hosting the competition!\n\nIs it permissible to post the link to the code in a public discussion here on Kaggle, however, the link would lead to a GitHub repository rather than a Kaggle notebook? I would like to share my code, but it is in a repository and is organized as multiple python scripts. (Currently, the repository is private.)", "2554982": "Yes, a paper will be written on the competition results.\n\n3 - The length of the sequences is 100 nucleotides. The leader, barcode, and tail will be added by the Das Lab.", "2555127": "Are you going to include authors of best solutions as authors of the paper? \n\nOur team will be happy to contribute to the paper", "2555260": "Yes, definitely OK to post on GitHub!", "2555263": "That's great to hear! \n\nAlso please note that there is no restriction on your team or other teams writing additional publications, if you have the bandwidth. \n\nLooking forward to more discussions as we put together publications.", "2555265": "Thank you!", "2555270": "rhijudas was that a 'yes'? 😄\n\nAnyway, +1 to @dmitrypenzar1996, I'd also appreciate if @rhijudas / @digitalembrace (or one of other organizers not present in this thread) could clarify whether the best solutions' authors would be able to co-author the flagship paper. Really excited to participate!", "2560446": "digitalembrace could you clarify this step? This also influence what we are going to investigate in our additional publication" }, "source": "meta" }