Paulina Szymczak
@szymczakpau.bsky.social
Computational biology by day, SciArt by night. PhD student at Ewa Szczurek's lab @Helmholtz Munich. Antimicrobial peptides & generative AI (she/her).
Reposted by Paulina Szymczak
If I have to explain to you why it's a bad idea to use a machine that makes shit up 15-80% of the time AND destroys the environment AND contributes to rising fascism and genocide AND puts brilliant skillful creatives out of work by stealing their work, then we're already speaking different languages
August 6, 2025 at 1:01 AM
If I have to explain to you why it's a bad idea to use a machine that makes shit up 15-80% of the time AND destroys the environment AND contributes to rising fascism and genocide AND puts brilliant skillful creatives out of work by stealing their work, then we're already speaking different languages
Reposted by Paulina Szymczak
Final call to join us the Munich Symposium on Generative AI in Molecule Discovery, for a day full of insights and networking opportunities, at our brand-new ✨ Conference Center @www.helmholtz-munich.de
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
June 18, 2025 at 8:54 AM
Final call to join us the Munich Symposium on Generative AI in Molecule Discovery, for a day full of insights and networking opportunities, at our brand-new ✨ Conference Center @www.helmholtz-munich.de
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
Reposted by Paulina Szymczak
Last days to register! Come and join us at the Symposium!
Final call to join us the Munich Symposium on Generative AI in Molecule Discovery, for a day full of insights and networking opportunities, at our brand-new ✨ Conference Center @www.helmholtz-munich.de
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
Check our agenda and secured your spot before June 24: events.hifis.net/event/2015/
June 18, 2025 at 11:49 AM
Last days to register! Come and join us at the Symposium!
Reposted by Paulina Szymczak
"Life is the shape it is for a purpose
And when you see how things really are
all the hurt and the waste falls away.
What's left is the beauty" Rosalind Franklin, in the "Life Story" movie (1987), especially recommended watching for early career scientists.
vimeo.com/179934156?&l...
And when you see how things really are
all the hurt and the waste falls away.
What's left is the beauty" Rosalind Franklin, in the "Life Story" movie (1987), especially recommended watching for early career scientists.
vimeo.com/179934156?&l...
May 25, 2025 at 11:19 AM
"Life is the shape it is for a purpose
And when you see how things really are
all the hurt and the waste falls away.
What's left is the beauty" Rosalind Franklin, in the "Life Story" movie (1987), especially recommended watching for early career scientists.
vimeo.com/179934156?&l...
And when you see how things really are
all the hurt and the waste falls away.
What's left is the beauty" Rosalind Franklin, in the "Life Story" movie (1987), especially recommended watching for early career scientists.
vimeo.com/179934156?&l...
Reposted by Paulina Szymczak
Dark academia: spending a lot of time in the microscopy room as a grad student
May 23, 2025 at 11:45 PM
Dark academia: spending a lot of time in the microscopy room as a grad student
Reposted by Paulina Szymczak
Save the date for the Helmholtz Munich AI for Health Symposium 2025, devoted to the topic of Generative AI in Molecule Discovery!
July 4, 2025
Helmholtz Munich Campus in Neuherberg, Germany
What now?
Register and submit abstracts (deadline May 2, 2025!)
events.hifis.net/event/2015/r...
July 4, 2025
Helmholtz Munich Campus in Neuherberg, Germany
What now?
Register and submit abstracts (deadline May 2, 2025!)
events.hifis.net/event/2015/r...
April 7, 2025 at 10:24 AM
Save the date for the Helmholtz Munich AI for Health Symposium 2025, devoted to the topic of Generative AI in Molecule Discovery!
July 4, 2025
Helmholtz Munich Campus in Neuherberg, Germany
What now?
Register and submit abstracts (deadline May 2, 2025!)
events.hifis.net/event/2015/r...
July 4, 2025
Helmholtz Munich Campus in Neuherberg, Germany
What now?
Register and submit abstracts (deadline May 2, 2025!)
events.hifis.net/event/2015/r...
Reposted by Paulina Szymczak
Use reward-guided generation and iterative refinement to steer the proteins generated from EvoDiff!
arxiv.org/abs/2502.14944
arxiv.org/abs/2502.14944
March 5, 2025 at 11:00 PM
Use reward-guided generation and iterative refinement to steer the proteins generated from EvoDiff!
arxiv.org/abs/2502.14944
arxiv.org/abs/2502.14944
Reposted by Paulina Szymczak
Programming languages named after people:
Ada, after Ada Lovelace
Haskell, after Haskell Curry
Bourne and Korn shells, after their inventors
Pike, after Rob Pike
Wolfram, after Stephen Wolfram
Julia, after no one in particular
Erlang, after Agner Erlang
Pascal, after Pedro Pascal
Ada, after Ada Lovelace
Haskell, after Haskell Curry
Bourne and Korn shells, after their inventors
Pike, after Rob Pike
Wolfram, after Stephen Wolfram
Julia, after no one in particular
Erlang, after Agner Erlang
Pascal, after Pedro Pascal
March 1, 2025 at 12:38 AM
Programming languages named after people:
Ada, after Ada Lovelace
Haskell, after Haskell Curry
Bourne and Korn shells, after their inventors
Pike, after Rob Pike
Wolfram, after Stephen Wolfram
Julia, after no one in particular
Erlang, after Agner Erlang
Pascal, after Pedro Pascal
Ada, after Ada Lovelace
Haskell, after Haskell Curry
Bourne and Korn shells, after their inventors
Pike, after Rob Pike
Wolfram, after Stephen Wolfram
Julia, after no one in particular
Erlang, after Agner Erlang
Pascal, after Pedro Pascal
Reposted by Paulina Szymczak
Hearing from Community Partners International, which works from Myanmar to Bangladesh, that USAID cuts mean 14,000 people with HIV losing access to lifesaving meds and a total stop in all HIV and TB screening. So many people will die due to these chaotic, sudden, horrifically implemented cuts.
February 28, 2025 at 4:48 PM
Hearing from Community Partners International, which works from Myanmar to Bangladesh, that USAID cuts mean 14,000 people with HIV losing access to lifesaving meds and a total stop in all HIV and TB screening. So many people will die due to these chaotic, sudden, horrifically implemented cuts.
Reposted by Paulina Szymczak
Here I react to the shameful encounter between American leaders and media and the Ukrainian president. No doubt there were more, but I saw five failures on our side: of hospitality, decency, democracy, strategy, and independence. Please watch and share.
snyder.substack.com/p/five-failu...
snyder.substack.com/p/five-failu...
Five Failures in the Oval Office (video)
America hurt itself badly today.
snyder.substack.com
February 28, 2025 at 10:15 PM
Here I react to the shameful encounter between American leaders and media and the Ukrainian president. No doubt there were more, but I saw five failures on our side: of hospitality, decency, democracy, strategy, and independence. Please watch and share.
snyder.substack.com/p/five-failu...
snyder.substack.com/p/five-failu...
Reposted by Paulina Szymczak
All probationary federal employees at NOAA’s EMC that’s responsible for keeping all US weather model systems running have been fired with 1 hours notice. And that includes me and colleagues. We will not go quietly because we care about the NOAA mission to protect the public.
February 27, 2025 at 9:18 PM
All probationary federal employees at NOAA’s EMC that’s responsible for keeping all US weather model systems running have been fired with 1 hours notice. And that includes me and colleagues. We will not go quietly because we care about the NOAA mission to protect the public.
Reposted by Paulina Szymczak
Per Dr Atul Gawande, ex USAID: “All malaria supplies protecting 53 million people, mostly children, including bed nets, diagnostics, preventive drugs, and treatments – terminated.”
There are no words that sufficiently describe this level of cruelty and sadism.
There are no words that sufficiently describe this level of cruelty and sadism.
February 28, 2025 at 5:16 AM
Per Dr Atul Gawande, ex USAID: “All malaria supplies protecting 53 million people, mostly children, including bed nets, diagnostics, preventive drugs, and treatments – terminated.”
There are no words that sufficiently describe this level of cruelty and sadism.
There are no words that sufficiently describe this level of cruelty and sadism.
Reposted by Paulina Szymczak
At the Antimicrobial peptides GRC Conference in beautiful Ventura, California. Thanks to Delphine Destoumieux-Garzón and Gerard Wood for the invitation and the opportunity to talk about Generative AI for antimicrobial peptide discovery! With
@szymczakpau.bsky.social, her poster and amazing sunset:)
@szymczakpau.bsky.social, her poster and amazing sunset:)
February 25, 2025 at 2:28 PM
At the Antimicrobial peptides GRC Conference in beautiful Ventura, California. Thanks to Delphine Destoumieux-Garzón and Gerard Wood for the invitation and the opportunity to talk about Generative AI for antimicrobial peptide discovery! With
@szymczakpau.bsky.social, her poster and amazing sunset:)
@szymczakpau.bsky.social, her poster and amazing sunset:)
Reposted by Paulina Szymczak
Good to see moscot-tools.org published in @nature.com ! We made existing Optimal Transport (OT) applications in single-cell genomics scalable and multimodal, added a novel spatiotemporal trajectory inference method and found exciting new biology in the pancreas! tinyurl.com/33zuwsep
Mapping cells through time and space with moscot - Nature
Moscot is an optimal transport approach that overcomes current limitations of similar methods to enable multimodal, scalable and consistent single-cell analyses of datasets across spatial and temporal...
tinyurl.com
January 23, 2025 at 8:42 AM
Good to see moscot-tools.org published in @nature.com ! We made existing Optimal Transport (OT) applications in single-cell genomics scalable and multimodal, added a novel spatiotemporal trajectory inference method and found exciting new biology in the pancreas! tinyurl.com/33zuwsep
feeling a but under the weather this week … thus an increased level of activity on social media and blog: kyunghyuncho.me/i-sensed-anx...
i sensed anxiety and frustration at NeurIPS’24 – Kyunghyun Cho
kyunghyuncho.me
December 22, 2024 at 8:01 PM
Reposted by Paulina Szymczak
1/🧬 Excited to share PLAID, our new approach for co-generating sequence and all-atom protein structures by sampling from the latent space of ESMFold. This requires only sequences during training, which unlocks more data and annotations:
bit.ly/plaid-proteins
🧵
bit.ly/plaid-proteins
🧵
December 6, 2024 at 5:44 PM
1/🧬 Excited to share PLAID, our new approach for co-generating sequence and all-atom protein structures by sampling from the latent space of ESMFold. This requires only sequences during training, which unlocks more data and annotations:
bit.ly/plaid-proteins
🧵
bit.ly/plaid-proteins
🧵
Reposted by Paulina Szymczak
Next up - contributed talk.
HelixFlow, SE(3)–equivariant Full-atom Design of Peptides With Flow-matching Models
Xuezhi Xie, Pedro A Valiente, Jisun Kim, Jin Sub Lee, Philip Kim
HelixFlow, SE(3)–equivariant Full-atom Design of Peptides With Flow-matching Models
Xuezhi Xie, Pedro A Valiente, Jisun Kim, Jin Sub Lee, Philip Kim
December 15, 2024 at 10:37 PM
Next up - contributed talk.
HelixFlow, SE(3)–equivariant Full-atom Design of Peptides With Flow-matching Models
Xuezhi Xie, Pedro A Valiente, Jisun Kim, Jin Sub Lee, Philip Kim
HelixFlow, SE(3)–equivariant Full-atom Design of Peptides With Flow-matching Models
Xuezhi Xie, Pedro A Valiente, Jisun Kim, Jin Sub Lee, Philip Kim
Reposted by Paulina Szymczak
Check out this bombshell publication by Michael Fischbach @mfgrp.bsky.social and coworkers @stanford-chemh.bsky.social @stanfordmedicine.bsky.social, imagine a cancer immune therapy/vaccine where treatment begins and ends with a painless topical skin treatment! 🤯
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Discovery and engineering of the antibody response to a prominent skin commensal - Nature
Nature - Discovery and engineering of the antibody response to a prominent skin commensal
www.nature.com
December 11, 2024 at 7:26 PM
Check out this bombshell publication by Michael Fischbach @mfgrp.bsky.social and coworkers @stanford-chemh.bsky.social @stanfordmedicine.bsky.social, imagine a cancer immune therapy/vaccine where treatment begins and ends with a painless topical skin treatment! 🤯
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Reposted by Paulina Szymczak
Hi @ddelalamo.bsky.social unfortunately, this paper from Jain et al. contains falsehoods, misleading comparisons, seemingly deliberate omissions, and is written in a tone not intended as a serious research paper. Please see our detailed response: www.linkedin.com/pulse/respon...
Response to Jain et al.
You may have seen a recent pre-print [1] from Jain et al. with strongly worded claims against the experimental results in our DiffDock paper [2].
www.linkedin.com
December 8, 2024 at 9:46 PM
Hi @ddelalamo.bsky.social unfortunately, this paper from Jain et al. contains falsehoods, misleading comparisons, seemingly deliberate omissions, and is written in a tone not intended as a serious research paper. Please see our detailed response: www.linkedin.com/pulse/respon...
Reposted by Paulina Szymczak
Looks like yet another case of overhyped results due to poor #bioMLeval evaluation of deep learning models -this time deep docking methods - specifically DiffDock. Look forward to the DiffDock authors response. But dont see any major flaws in this critique. Conclusion is REALLY worth reading!
Been a while since I read a paper like this:
• "What [DiffDock] appears to be doing cannot be considered" docking
• "Results are ... contaminated with near neighbors to test cases"
• "Results for DiffDock were artifactual"
• "Results for other methods were incorrectly done"
arxiv.org/abs/2412.02889
• "What [DiffDock] appears to be doing cannot be considered" docking
• "Results are ... contaminated with near neighbors to test cases"
• "Results for DiffDock were artifactual"
• "Results for other methods were incorrectly done"
arxiv.org/abs/2412.02889
Deep-Learning Based Docking Methods: Fair Comparisons to Conventional Docking Workflows
The diffusion learning method, DiffDock, for docking small-molecule ligands into protein binding sites was recently introduced. Results included comparisons to more conventional docking approaches, wi...
arxiv.org
December 5, 2024 at 11:30 PM
Looks like yet another case of overhyped results due to poor #bioMLeval evaluation of deep learning models -this time deep docking methods - specifically DiffDock. Look forward to the DiffDock authors response. But dont see any major flaws in this critique. Conclusion is REALLY worth reading!
Reposted by Paulina Szymczak
Oh wow. Maybe kind of a harsh attack, but we really do need a healthy debate about docking benchmarks. While I'm excited to see this one being fought out, may I already suggest the next point of debate: can we find a better metric than the utterly inadequate 2 Å RMSD cutoff?
Been a while since I read a paper like this:
• "What [DiffDock] appears to be doing cannot be considered" docking
• "Results are ... contaminated with near neighbors to test cases"
• "Results for DiffDock were artifactual"
• "Results for other methods were incorrectly done"
arxiv.org/abs/2412.02889
• "What [DiffDock] appears to be doing cannot be considered" docking
• "Results are ... contaminated with near neighbors to test cases"
• "Results for DiffDock were artifactual"
• "Results for other methods were incorrectly done"
arxiv.org/abs/2412.02889
Deep-Learning Based Docking Methods: Fair Comparisons to Conventional Docking Workflows
The diffusion learning method, DiffDock, for docking small-molecule ligands into protein binding sites was recently introduced. Results included comparisons to more conventional docking approaches, wi...
arxiv.org
December 5, 2024 at 4:13 PM
Oh wow. Maybe kind of a harsh attack, but we really do need a healthy debate about docking benchmarks. While I'm excited to see this one being fought out, may I already suggest the next point of debate: can we find a better metric than the utterly inadequate 2 Å RMSD cutoff?
Reposted by Paulina Szymczak
scRNA-seq has revolutionized biology, but it’s not without challenges. We uncovered significant quality issues in widely used reference cell atlases like the Human Cell Atlas and Tabula Sapiens. Up to 85% of cells in some datasets are low-quality or misidentified!
2/8
2/8
December 3, 2024 at 8:38 AM
scRNA-seq has revolutionized biology, but it’s not without challenges. We uncovered significant quality issues in widely used reference cell atlases like the Human Cell Atlas and Tabula Sapiens. Up to 85% of cells in some datasets are low-quality or misidentified!
2/8
2/8
Reposted by Paulina Szymczak
Excited to share our study led by @SakenovaNazgul! We used chemical genetics to identify and understand cross-resistance & collateral-sensitivity (CS) between antibiotics and reduced AMR development with CS drug pairs @embl.org with @camille_goemans @EPFL_en www.nature.com/articles/s41...
Systematic mapping of antibiotic cross-resistance and collateral sensitivity with chemical genetics - Nature Microbiology
Resistance to one antibiotic can make bacteria resistant or sensitive to another antibiotic, opening paths for combinatorial treatments. This study presents an approach to systematically discover and ...
www.nature.com
December 2, 2024 at 10:33 AM
Excited to share our study led by @SakenovaNazgul! We used chemical genetics to identify and understand cross-resistance & collateral-sensitivity (CS) between antibiotics and reduced AMR development with CS drug pairs @embl.org with @camille_goemans @EPFL_en www.nature.com/articles/s41...
Reposted by Paulina Szymczak
I have created a starter pack for the Polish ML community. Let's grow it together! Let me know if you know somebody who should be included.
go.bsky.app/TSkgFjK
go.bsky.app/TSkgFjK
November 21, 2024 at 8:13 AM
I have created a starter pack for the Polish ML community. Let's grow it together! Let me know if you know somebody who should be included.
go.bsky.app/TSkgFjK
go.bsky.app/TSkgFjK