Going into @workshopmlsb.bsky.social , a few thoughts on the PLINDER dataset. This is a monumental and highly appreciated undertaking for extracting protein-ligand interactions from the PDB. First the good...
December 15, 2024 at 6:52 PM
Going into @workshopmlsb.bsky.social , a few thoughts on the PLINDER dataset. This is a monumental and highly appreciated undertaking for extracting protein-ligand interactions from the PDB. First the good...
We’re pre-printing this early to get feedback from the community. We teamed up with @caswognum.nl at @polarishub.io to make the dataset and benchmark ML-ready polarishub.io/datasets/pli.... See also our github.com/plinder-org/... for more. Would love to hear your thoughts! (8/n
runs-n-poses-dataset
plinder-org
polarishub.io
February 8, 2025 at 10:20 AM
We’re pre-printing this early to get feedback from the community. We teamed up with @caswognum.nl at @polarishub.io to make the dataset and benchmark ML-ready polarishub.io/datasets/pli.... See also our github.com/plinder-org/... for more. Would love to hear your thoughts! (8/n
Here's a year-end update from #PLINDER. It's been really great working on this project and all the other projects that it has kickstarted - the gift that keeps on giving.
Happy holidays everyone!
www.plinder.sh/blog/updates
Happy holidays everyone!
www.plinder.sh/blog/updates
Updates
Updates
www.plinder.sh
December 24, 2024 at 3:56 PM
Here's a year-end update from #PLINDER. It's been really great working on this project and all the other projects that it has kickstarted - the gift that keeps on giving.
Happy holidays everyone!
www.plinder.sh/blog/updates
Happy holidays everyone!
www.plinder.sh/blog/updates
Re docking: @workshopmlsb.bsky.social has been piloting a leaderboard for the PLINDER benchmark. It would be awesome to get submissions using existing methods, then everyone can see comparisons better 😁
leaderboard & instructions are on mlsb.io
leaderboard & instructions are on mlsb.io
December 9, 2024 at 4:15 PM
Re docking: @workshopmlsb.bsky.social has been piloting a leaderboard for the PLINDER benchmark. It would be awesome to get submissions using existing methods, then everyone can see comparisons better 😁
leaderboard & instructions are on mlsb.io
leaderboard & instructions are on mlsb.io
Keel ja Kirjandus: kust on tulnud sõnad “obadus” ja “plinder”, kirjutab Lembit Vaba. doi.org/10.54013/kk8...
January 6, 2026 at 10:49 AM
Keel ja Kirjandus: kust on tulnud sõnad “obadus” ja “plinder”, kirjutab Lembit Vaba. doi.org/10.54013/kk8...
Hearing that kier starmer has blayed a plinder
April 20, 2026 at 3:23 PM
Hearing that kier starmer has blayed a plinder
I want to thank my co-authors @jeeberhardt.bsky.social, @torstenschwede.bsky.social, @ninjani.bsky.social and all of our collaborators! RunsN’ Poses builds on PLINDER and OpenStructure—this work wouldn’t be possible without them!
Also thanks to @rokbreznikar.bsky.social for this amazing logo! 9/9
Also thanks to @rokbreznikar.bsky.social for this amazing logo! 9/9
February 8, 2025 at 10:37 AM
I want to thank my co-authors @jeeberhardt.bsky.social, @torstenschwede.bsky.social, @ninjani.bsky.social and all of our collaborators! RunsN’ Poses builds on PLINDER and OpenStructure—this work wouldn’t be possible without them!
Also thanks to @rokbreznikar.bsky.social for this amazing logo! 9/9
Also thanks to @rokbreznikar.bsky.social for this amazing logo! 9/9
Re: quality, partly a docs issue - "system_pass_validation_criteria" combines github.com/plinder-org/.... But still also a standards issue - 1) ~1/3 of the data passes, rest has useful info, 2) cryo-EM elephant in the room, 3) working to add density fit into our metrics. Suggestions welcome! (4/n)
plinder/src/plinder/data/utils/annotations/aggregate_annotations.py at 9658cc294722c164f15241f478c4e612cfa437d8 · plinder-org/plinder
Protein Ligand INteraction Dataset and Evaluation Resource - plinder-org/plinder
github.com
December 16, 2024 at 8:48 AM
Re: quality, partly a docs issue - "system_pass_validation_criteria" combines github.com/plinder-org/.... But still also a standards issue - 1) ~1/3 of the data passes, rest has useful info, 2) cryo-EM elephant in the room, 3) working to add density fit into our metrics. Suggestions welcome! (4/n)
@peterskrinjar.bsky.social presenting #CAMEO and #PLINDER at #CASP16 , i.e what we at the @torstenschwede.bsky.social group have been up to since last CASP in automating benchmarking of complex prediction.
December 1, 2024 at 11:04 PM
@peterskrinjar.bsky.social presenting #CAMEO and #PLINDER at #CASP16 , i.e what we at the @torstenschwede.bsky.social group have been up to since last CASP in automating benchmarking of complex prediction.
Next up we're highlighting the results of the PINDER and PLINDER challenge.
December 15, 2024 at 11:07 PM
Next up we're highlighting the results of the PINDER and PLINDER challenge.
End of year notes from #PLINDER team!
It's been an incredible year for datasets and benchmarks for the #ML / #CompChem / #CompBio community.
Superbly grateful for the positive responses, constructive feedback and contributions!
It's just the start, lots more in 2️⃣0️⃣2️⃣5️⃣ 🎉
www.plinder.sh/blog/updates
It's been an incredible year for datasets and benchmarks for the #ML / #CompChem / #CompBio community.
Superbly grateful for the positive responses, constructive feedback and contributions!
It's just the start, lots more in 2️⃣0️⃣2️⃣5️⃣ 🎉
www.plinder.sh/blog/updates
Updates
Updates
www.plinder.sh
December 24, 2024 at 3:04 PM
End of year notes from #PLINDER team!
It's been an incredible year for datasets and benchmarks for the #ML / #CompChem / #CompBio community.
Superbly grateful for the positive responses, constructive feedback and contributions!
It's just the start, lots more in 2️⃣0️⃣2️⃣5️⃣ 🎉
www.plinder.sh/blog/updates
It's been an incredible year for datasets and benchmarks for the #ML / #CompChem / #CompBio community.
Superbly grateful for the positive responses, constructive feedback and contributions!
It's just the start, lots more in 2️⃣0️⃣2️⃣5️⃣ 🎉
www.plinder.sh/blog/updates
Inspired by Plinder, the team at Bioptic (bioptic.io) released a new binding affinity dataset called Plumber on Polaris: polarishub.io/datasets/opt...
From KiB to TiB and from small molecules to phenomics, the Polaris Hub is built to support any ML-ready dataset in drug discovery.
Learn how! 🧵
From KiB to TiB and from small molecules to phenomics, the Polaris Hub is built to support any ML-ready dataset in drug discovery.
Learn how! 🧵
plumber
optic
polarishub.io
November 20, 2024 at 12:10 AM
Inspired by Plinder, the team at Bioptic (bioptic.io) released a new binding affinity dataset called Plumber on Polaris: polarishub.io/datasets/opt...
From KiB to TiB and from small molecules to phenomics, the Polaris Hub is built to support any ML-ready dataset in drug discovery.
Learn how! 🧵
From KiB to TiB and from small molecules to phenomics, the Polaris Hub is built to support any ML-ready dataset in drug discovery.
Learn how! 🧵
PLINDER has a single test set that is separated from the training set by removing all systems that are too similar to the test set. We've been looking at descriptors of pocket similarity that are distinct to those used in the construction of the PLINDER test/train split.
December 15, 2024 at 6:52 PM
PLINDER has a single test set that is separated from the training set by removing all systems that are too similar to the test set. We've been looking at descriptors of pocket similarity that are distinct to those used in the construction of the PLINDER test/train split.
= Going through some abandoned villages =
Me: Here we go a-plundering
Jas: Heck yeah!!
Me: What do people normally say? Plunder and?
Jas: Plinder
Me: What no Jas
Jas: Plinder plander ploonder 🎵
Me: Fucking heck
[ 23 September 2021 ]
Me: Here we go a-plundering
Jas: Heck yeah!!
Me: What do people normally say? Plunder and?
Jas: Plinder
Me: What no Jas
Jas: Plinder plander ploonder 🎵
Me: Fucking heck
[ 23 September 2021 ]
November 12, 2024 at 6:00 AM
= Going through some abandoned villages =
Me: Here we go a-plundering
Jas: Heck yeah!!
Me: What do people normally say? Plunder and?
Jas: Plinder
Me: What no Jas
Jas: Plinder plander ploonder 🎵
Me: Fucking heck
[ 23 September 2021 ]
Me: Here we go a-plundering
Jas: Heck yeah!!
Me: What do people normally say? Plunder and?
Jas: Plinder
Me: What no Jas
Jas: Plinder plander ploonder 🎵
Me: Fucking heck
[ 23 September 2021 ]
Now onto the less good... When @alutsky.bsky.social trained a binding affinity prediction model on PLINDER, he got a negative correlation, which was surprising looking at the predictions (note only 182 test set systems having an affinity annotation).
December 15, 2024 at 6:52 PM
Now onto the less good... When @alutsky.bsky.social trained a binding affinity prediction model on PLINDER, he got a negative correlation, which was surprising looking at the predictions (note only 182 test set systems having an affinity annotation).
These ligands make up 33% of the v2 plinder test set (shown with counts):
MG 94
NAG 72
NAG-NAG 64
ZN 44
MN 32
BMP 30
HEM 24
EDO 18
SO4 17
PO4 16
ADP 15
GOL 13
ANP 13
CA 13
FMN 11
MG 94
NAG 72
NAG-NAG 64
ZN 44
MN 32
BMP 30
HEM 24
EDO 18
SO4 17
PO4 16
ADP 15
GOL 13
ANP 13
CA 13
FMN 11
Here's a year-end update from #PLINDER. It's been really great working on this project and all the other projects that it has kickstarted - the gift that keeps on giving.
Happy holidays everyone!
www.plinder.sh/blog/updates
Happy holidays everyone!
www.plinder.sh/blog/updates
Updates
Updates
www.plinder.sh
December 24, 2024 at 5:06 PM
These ligands make up 33% of the v2 plinder test set (shown with counts):
MG 94
NAG 72
NAG-NAG 64
ZN 44
MN 32
BMP 30
HEM 24
EDO 18
SO4 17
PO4 16
ADP 15
GOL 13
ANP 13
CA 13
FMN 11
MG 94
NAG 72
NAG-NAG 64
ZN 44
MN 32
BMP 30
HEM 24
EDO 18
SO4 17
PO4 16
ADP 15
GOL 13
ANP 13
CA 13
FMN 11
New benchmarks like plinder and models like AlphaFold3 now introduced the pli-lddt metric, which has a dynamic threshold similar to auroc.
www.plinder.sh
www.plinder.sh
PLINDER: Unlocking new discoveries through a scale-shift in data and real-world relevant and rigorous evaluations
PLINDER is an academic-industry collaboration offering a gold standard dataset and evaluations for computational protein-ligand interactions prediction.
www.plinder.sh
December 6, 2024 at 1:11 PM
New benchmarks like plinder and models like AlphaFold3 now introduced the pli-lddt metric, which has a dynamic threshold similar to auroc.
www.plinder.sh
www.plinder.sh
With @polarishub.io we’ve built out infrastructure to make hosting and using large scale, ML ready dataset and benchmarks easy. In addition to the dataloaders, we could also maintain leaderboards.
Would be happy to explore ways in which we can support the awesome work you’re doing with Plinder.
Would be happy to explore ways in which we can support the awesome work you’re doing with Plinder.
December 17, 2024 at 12:24 AM
With @polarishub.io we’ve built out infrastructure to make hosting and using large scale, ML ready dataset and benchmarks easy. In addition to the dataloaders, we could also maintain leaderboards.
Would be happy to explore ways in which we can support the awesome work you’re doing with Plinder.
Would be happy to explore ways in which we can support the awesome work you’re doing with Plinder.
i'm blaying a plinder now. full schedule
January 18, 2026 at 1:01 AM
i'm blaying a plinder now. full schedule
Plunder, dammit. Why even suggest 'Plinder' autocorrect?
July 2, 2026 at 10:16 PM
Plunder, dammit. Why even suggest 'Plinder' autocorrect?
If the test set is good, I am less concerned about the train set - let the developer figure out what works best - but I don’t want the bold numbers on the PLINDER benchmark to go to glycosylation predictors.
December 16, 2024 at 8:25 PM
If the test set is good, I am less concerned about the train set - let the developer figure out what works best - but I don’t want the bold numbers on the PLINDER benchmark to go to glycosylation predictors.
This highlights what I think is the most critical correction needed to the PLINDER test set: getting the right distribution. Obviously, 16% of the test set should not be the exact same ligand, but we should also think more holistically about what a good test distribution should look like.
December 15, 2024 at 6:52 PM
This highlights what I think is the most critical correction needed to the PLINDER test set: getting the right distribution. Obviously, 16% of the test set should not be the exact same ligand, but we should also think more holistically about what a good test distribution should look like.
Using PLINDER train/test uncritically for protein-ligand training is unlikely to end well (but easy enough to fix with stratification and quality filtering).
December 15, 2024 at 9:39 PM
Using PLINDER train/test uncritically for protein-ligand training is unlikely to end well (but easy enough to fix with stratification and quality filtering).
The systems close to the test set have been removed, although it is possible PLINDER is being overly conservative as quite a few systems are removed that, according to these descriptors, are not close to the test set.
December 15, 2024 at 6:52 PM
The systems close to the test set have been removed, although it is possible PLINDER is being overly conservative as quite a few systems are removed that, according to these descriptors, are not close to the test set.