CoronaWhy

is an international group of 900+ volunteers whose mission is to improve global coordination on COVID-19

[CoronaWhy @CoronaWhy]

Really interesting thinking about how to keep efforts sustainable - something we’ve been introspecting on a lot

[Gyori Lab for Computational Biomedicine @gyorilab]

Our @benjamingyori is presenting today on automated hypothesis generation and human-machine collaboration at the 1st Workshop on Nobel Turing Challenge: https://t.co/B9TxemiOqa

[Gyori Lab for Computational Biomedicine @gyorilab]

The @Bioregistry has an API that links entities to external resolvers and providers like in @OBOFoundry @IdentifiersOrg @EBIOLS:

🧬 TP53 https://t.co/YmcczeyebW
🧪 Aspirin https://t.co/qxgBb3IphM
🐕 Dog https://t.co/G3y5HbiGMO
🦠 COVID-19 https://t.co/xKz4m72LM1

[PyKEEN @keenuniverse]

PyKEEN v1.8.0 is out with three huge new features:

🔥 Inductive link prediction and the NodePiece model
📐 Novel, comparable z-scored rank-based metrics
💾 Under-the-hood improvement to negative sampiling

Full change log at https://t.co/C83KmaSWYU

[PyKEEN @keenuniverse]

🚨 We're happy to share one of our submissions to @GLB_Workshop outlining new, more comparable link prediction metrics than MR, MRR, and Hits@K including the zMR, zMRR, and zHits@K

Thanks to authors @cthoyt @BerrendorfMax @michael_galkin @vtresp @benjamingyori

[Gyori Lab for Computational Biomedicine @gyorilab]

🚨 Our @Bioregistry project imports and standardizes prefixes/metadata for biomedical nomenclatures from 15 external registries (e.g. @OBOFoundry @IdentifiersOrg @bioportal) and allows for community 👪 extension using an open data model at https://t.co/rTk7RwtMo8.

[Kevin M Jablonka @kmjablonka]

Our speaker's list for our @cecamEvents workshop (/w @JoshuaSchrier @BenBlaiszik @petermurrayrust @helsoeste @cthoyt) is now finalized! Consider registering if you're interested in making chemical data machine-actionable! @EPFLOpenScience #chemtwitter

[Gyori Lab for Computational Biomedicine @gyorilab]

Our @benjamingyori gave the first @cosi_sysmod webinar for @iscb Academy describing our team's work.

📽️ Full recording: https://t.co/Q0kaZNYica

🔗 More about what we do: https://t.co/9GaO8FtU1g

[PyKEEN @keenuniverse]

🚨 PyKEEN 1.7.0 is out! 🚨

New: 7 models, 3 encoders, 2 trackers, keras-style training callbacks, uncertainty estimation, and more

📖 Full Release Notes: https://t.co/oaYAAix2bL

💪 Huge S/O to external contributors @dobraczka @michael_galkin @SBonner0 @Ralph_Abb and others

[Int Soc Biocuration @biocurator]

ISB2022 Biocuration Conference first session 🎉
🗓️ 29 March 2022
➡️ Panel: Promotion and dissemination of biocuration efforts
➡️ Invited Speakers Talks
Submit your abstracts: https://t.co/i2IIFV2E6S
More info: https://t.co/hoTUy9ZUGW

[SysMod @cosi_sysmod]

📣The first SysMod Webinar is coming up. Dr. Gyori (@benjamingyori from @HavardUnivers) will talk about “Accelerating biomedical discovery with large-scale knowledge assembly and human-machine collaboration” – join us on January 18, 2022, at 11:00 AM EST: https://t.co/8JsRX4wFRZ

[Gyori Lab for Computational Biomedicine @gyorilab]

We've been developing PyBioPAX, a native Python object model for BioPAX Level 3 with parsing, serializing, and traversal of BioPAX models.

🚀 Install: pip install pybiopax

🖥️ Code: https://t.co/oDcUirobc9

📖 Docs: https://t.co/YAZ0jcZYl1

[Gyori Lab for Computational Biomedicine @gyorilab]

We’re still actively maintaining FamPlex, a resource for text mining and reasoning over human protein families/complexes. Here's some links to learn more and a thread on a few improvements we made recently:

🖥️Code/data: https://t.co/vTZHYr8NB1

📜Paper: https://t.co/3NaecZtIuL

[PyKEEN @keenuniverse]

PyKEEN 1.6.0 is out, now with batteries from PyTorch 1.9 built in!

🎁 It’s got new models, new datasets, new loss functions, bug fixes, and improvements.

🙏 Huge s/o to our many external contributors for this release

📖 Release notes: https://t.co/3LQ3DRZuO0

[CoronaWhy @CoronaWhy]

Our team has built a prototype of a literature review engine for efficient navigation of COVID-19-focused medical literature. Interested in participating in a user testing session with us to test out our prototype? Please fill out this survey! #epitwitter https://t.co/65KJqH2O7O

[CoronaWhy @CoronaWhy]

The next generation of a tool developed and used by our members to generate novel biological hypotheses using knowledge graphs desscribing COVID-19! Congrats!

[PyKEEN @keenuniverse]

PyKEEN v1.0.4 is out with improved HPO and the new constrained evaluation feature! Get it with:

pip install -U pykeen

and check the release notes at https://t.co/BEHUdPYzS1

[CoronaWhy @CoronaWhy]

Perfect timing - today our Vaccines and Therapeutics team will begin its work on link prediction on our networks in Biological Expression Language networks using PyKEEN!

[SDA Research @SDA_Research]

We are thrilled to share our pre-print entitled "Bringing Light Into the Dark: A Large-scale Evaluation of Knowledge Graph Embedding Models Under a Unified Framework” https://t.co/6qvjUdI31x #knowledgegraphs #machinelearning #benchmarking #AcademicChatter

[Slava Tykhonov 🇳🇱🇺🇦 @4tykhonov]

@CoronaWhy is building distributed and scaled #Covid_19 infra for @openscience suitable for other tasks like cancer research. It's Operating System with #AI and public ML models created to build #coronavirus Knowledge Graph for Social and Life Sciences https://t.co/R20w3a66tC

[PyBEL @pybelbio]

A quick overview by Dan Sosa at the Boston Data Science Meetup on @CoronaWhy’s plans for generation of knowledge graphs in the Biological Expression Language (BEL) and later structural causal models to support counterfactual reasoning

https://t.co/4eciCJO7oi

[CoronaWhy @CoronaWhy]

A recording of @cthoyt’s talk “Reusable, Reproducible, Useful Computational Science in Python” is now available on YouTube at https://t.co/EExstvLnvg