If youโre interested in the current landscape of federated learning in biomedical research, you can check the paper here:
๐ doi.org/10.1093/narg...
@mariachdiez
Systems engineer turned bioinformatician ๐ฉ๐ปโ๐ป Biomedical Sciences PhD at @bsc-cns.bsky.social and @ub.edu. Open Science ๐ฑ useGalaxy โจ Federated Learning ๐งฎ Also ๐ฎ๐๐๐ปโโ๏ธ๐ง๐ป๐ธ
If youโre interested in the current landscape of federated learning in biomedical research, you can check the paper here:
๐ doi.org/10.1093/narg...
I'm really excited to call this work my first first author paper, but this project was truly a team effort ๐ I would especially like to thank Carles Hernandez-Ferrer, Laia Codรณ, Josep Lluis Gelpi and Salvador Capella-Gutierrez for their support and constructive feedback throughout this journey โค๏ธ
This analysis highlights the need to evolve toward modular software architectures that enable greater interoperability and facilitate rapid adoption of new regulatory scenarios, supporting more advanced privacy-preserving techniques while still serving researchers' needs.
Our findings suggest that many of the frameworks were initially developed for specific use cases and later extended to other applications, often with a focus on enabling the use of data across silos, rather than across devices, and on horizontal architecture for sharing the same data type.
Federated learning frameworks allow machine-learning models to be trained on external, sensitive data without compromising security or privacy ๐ In this work, we systematically reviewed existing frameworks to evaluate their long-term sustainability, flexibility, and usability in biomedical research.
๐ฃ After over a year of hard work, revisions, and collaboration, Iโm excited to share that our paper, โFederated learning frameworks: quality and interoperability for biomedical research,โ is now published in NAR Genomics and Bioinformatics ๐ Mini thread incoming!
doi.org/10.1093/narg...
Two open positions for #EUCAIM project (sorry for the crossposting)
x.com/INB_Official...
Los datos sobre los premios Nobel revelan una red que se resume en tres consejos si quiere aspirar a uno:
โEncuentre a alguien que ya lo haya ganado y trabaje con รฉl, procure que su familia estรฉ en el 5% de las que mรกs dinero ganan y no sea mujerโ.
Artรญculo + ๐งต
dixitciencia.com/2024/11/12/l...
A photo of Jane Goodall, in her later days with a quote from her: "The greatest danger to our future is our apathy. Each one of us must take responsibility for our own lives, and above all, show respect and love for living things around us, especially each other."
RIP Jane Goodall, who has died at 91.
She was a gentle disrupter who reshaped how we understand our place in nature. Her research with chimpanzees revolutionized science, and her boundless optimism touched the world.
buff.ly/5Hwx3Z7 ๐๐ฉบ
Incredibly happy to have been able to present our work even though it couldn't be in person. Hoping to be there with more next time!!
Scientific productivity gap based on English-language peer-reviewed papers. Shown are the maximum % differences in the number of peer-reviewed papers published by female native English speakers from a high-income country (-45%), female non-native English speakers from a high-income country (-60%), and female non-native English speakers from a lower-middle income country (-70%), compared to male native English speakers from a high-income country (red flag). Credit: Tatsuya Amano
How do your #linguistic, #economic & #gender backgrounds impact your #scientific productivity? @tatsuya-amano.bsky.social & co reveal that being a woman, a non-native English speaker, and from a low-income country is associated with a 70% reduction in productivity @plosbiology.org ๐งช plos.io/4n3RLRQ
Academic authors, here's a peek into the black box of journal publishing from an journal editor if you can bear it:
๐จ New in PNAS!
๐งฌ 64% of disease co-occurrences can be explained by transcriptomic similarities.
Comorbidities arenโt randomโthey have a molecular basis.
Hereโs how we found it ๐ (1/n)
๐ doi.org/10.1073/pnas.2421060122
@alfonsovalencia.bsky.social
Missed GBCC2025? Want to rewatch your favorite talks?
All session recordings (including Galaxy Live!) are now available on our YouTube channel. Revisit the presentations, explore new tools and workflows, and catch up on sessions you may have missed.
Watch here: youtube.com/playlist?lis...
Iโve been noticing this for a while. We call for multidisciplinary teams, yet infrastructure and computation are often treated as a separate world from pure experimental biology. We need stronger integration across these domains for projects (and science!) to move forward.
Really cool read if you're into federated analysis in genomics ๐งฌ It clears up a lot of the jargon and makes it easier to get the bigger picture! rdcu.be/eyKGr
Wish I could have been there to see your talks! ๐
The "reproducibility crisis" in science constantly makes headlines. Repro efforts are often limited. What if you could assess reproducibility of an entire field?
That's what @brunolemaitre.bsky.social et al. have done. Fly immunity is highly replicable & offers lessons for #metascience
A ๐งต 1/n
From student to researcher, a #career in #science can come with a high price tag. @drcraigmc.bsky.social โฌexplores how wealth shapes opportunity in #STEM and proposes structural changes to support #equity and inclusion. ๐งช
plos.io/4edGlY4
Iโve said it before, and Iโll say it again: We refuse to reinforce systems of oppression and exclusion.
If you agree, take my grandmotherโs advice. Put your money where your mouth is ๐ซถ๐พ
Donate: www.zeffy.com/en-US/donati...
New research alert ๐โโโ
In our latest pre-print, 2nd of my PhD, we performed a Systematic Literature Review on the use of Deep Representational Learning (DRL), especially the Variational Autoencoder (VAE), in cancer progression research.
This thread explains our main findings.
(1 minute read)
Today is Day 2 of Curating the Clinical Genome organised by @deciphergenomics.bsky.social and hosted by @eventswcs.bsky.social ๐ซถ
Come visit poster 15 - I would love to talk to you about how the fabulous members of the @ga4gh.org community are accelerating clinical genomics! ๐ฑ
#CCG2025
๐ 2025 ELIXIR All Hands Meeting #ELIXIR25 has come to an end! ๐ Thank you all for joining us in Thessaloniki ๐ฌ๐ท and we look forward to seeing you again next year in Lyon, France ๐ซ๐ท for #ELIXIR26! @ifb-elixir-fr.bsky.social
If you're interested in patterns across FL frameworks, and how real-world federated learning use cases align with them, come check out Poster 76 at tomorrowโs poster session! #ELIXIR25
Excited to share our new preprint: The impact of sex, age, and genetic ancestry on DNA methylation across tissues. Briefly, we explored how sex, age, and genetic ancestry shape DNA methylation across tissues and individuals. Want to know a bit more? Follow this โbluetorialโ!
doi.org/10.1101/2025...
The Edinburgh Open Research Conference is always a great event to go along to. Itโs free and hybrid, and registration is open now! #EORC25
GitHub reports Python now leads in popularity, fueled by generative AI. For new researchers, language choice depends on the task: compiled (C++, Rust) for tools, scripting (Python, R) for data. #AcademicSky #Programing ๐งฌ๐ป
1. LLM-generated code tries to run code from online software packages. Which is normal but
2. The packages donโt exist. Which would normally cause an error but
3. Nefarious people have made malware under the package names that LLMs make up most often. So
4. Now the LLM code points to malware.
Congratulations Winona!! ๐๐
A week left to prepare for the project proposal for the #BioHackEU25.
๐งโ๐ป Join hundreds of willing #bioinformaticians
๐งฌ Huge range of expertise in #bioinformatics
๐ฅ Work on blue sky ideas or advanced existing projects
๐Submission deadline 14 April: loom.ly/5gopDeQ