justaddcoffee

Justin Reese, PhD | computational biology | machine learning | biomedical data | knowledge graphs | running

[mhackeroni @mhackeroni]

This guy is the reason why we played Hack-a-Sat to begin with. Thanks @chqmatteo 🙏

[Chris Mungall @chrismungall]

Here's our pre-print describing our GPT-3 based knowledge extraction tool SPIRES: https://t.co/8b57yidudx. Great work from @harry_caufield et al! SPIRES allows you to specify a knowledge schema (in @linkml_data) and then populate instances of that schema from unstructured text

[justaddcoffee @justaddcoffee]

@ontowonka @MonarchInit @hp_ontology @AMIAinformatics Many thanks @ontowonka for all your help making this project possible!

[Peter Robinson @pnrobins]

A pleasure to collaborate with Elena Casiraghi on so many cool projects! Here's the latest:
A method for comparing multiple imputation techniques: a case study on the U.S. National COVID Cohort Collaborative
https://t.co/O4w5Q75oZB

[Cmckenna @McKennaGenetics]

If you work in a Clinical Genetics department that recruited to the 100k Genome Project I *highly* recommend encouraging one of your clinical genetics trainees to review the top 5 exomiser results in your undiagnosed patients (particularly the trios).

[justaddcoffee @justaddcoffee]

Thanks @SecGranholm for sharing our work!

[Secretary Jennifer Granholm @SecGranholm]

This crucial research will help fill the gaps in our understanding of how #LongCOVID works and develop more effective treatments for those still suffering. https://t.co/bhYvMNEvav

[justaddcoffee @justaddcoffee]

Interesting statistical problem I hadn't heard of:
https://t.co/l86ruT7bma

[GRAPE @GRAPElib]

Woohoo! #GRAPE just hit 100 stars on GitHub! Thank you to all the amazing developers who have supported our graph representation learning library. We couldn't have done it without you! 🍇💜🍇 #opensource #machinelearning

[Luca Cappelletti @LucaCappellett6]

Scientific #publications are a treasure trove, but integrating their knowledge into #graph #machinelearning is more challenging than it needs to be! 🧵 1/4

[GRAPE @GRAPElib]

Pushing the ✉️ of 🍇's @psresnik score implementation by computing 3T, i.e. 3*10^12, scores from @NCBI Taxonomy (2438821 nodes) upper triangular matrix.

This is heavily parallelized and takes ≈3h on a 💻 with 8GBs of RAM and 96 cores.

[justaddcoffee @justaddcoffee]

Our preprint demonstrating the association of metformin with reduced COVID-19 severity in patients with prediabetes:

https://t.co/syVblHukli

[GRAPE @GRAPElib]

#GraphML 🍇 preprint is on @arxiv!

❤️ to @LucaCappellett6, @Zommiommy, Casiraghi, @RavanmehrVida, @Tiff_callahan, @realmarcin, @chrismungall, @pnrobins, @justaddcoffee & Valentini

📍@MonarchInit, @ConsorzioCINI, @LaStatale, @jacksonlab & @BerkeleyLab

📜:https://t.co/LmVy0Nr5K6

[GRAPE @GRAPElib]

Animations >> Words!

Here's @justaddcoffee's KGCOVID19 embedding (~574k nodes, ~18M edges), animated in ~4m on my desktop using 🍇's @rustlang implementation (with #Python bindings) of @tangjianpku's First-order LINE.

Use it with @googleslides!

💻: https://t.co/SLgBmuhlqg

[GRAPE @GRAPElib]

The new version of 🍇, 0.1.10, is out! We introduce support for 14 MMAP-ed graph embedding algorithms, plus a significantly extended set of tutorials!

E.g. TSNEs of 36 node embeddings of Cora, find the code here: https://t.co/iX6rmHXKcJ

[justaddcoffee @justaddcoffee]

Our latest paper identifying subtypes of #longcovid using machine learning on EHR data from the National COVID Cohort Collaborative (N3C):

https://t.co/jWw7OkSDTS

[Nicolas Matentzoglu @NicoMatentzoglu]

After a lot of hard work by many people it’s finally done: Simple Standard for Sharing Ontological Mappings (SSSOM). Thank you to all contributors! https://t.co/9QGvrFd7ZL

[Peter Robinson @pnrobins]

An algorithmic framework for isoform-specific functional analysis https://t.co/fOWKFK6F6y @MonarchInit

[Chris Mungall @chrismungall]

Biolink Model is our proposed standard for biological #KnowledgeGraphs, created by the @ncats_nih_gov #BiomedicalDataTranslator team https://t.co/yciV6wDjdY

[justaddcoffee @justaddcoffee]

https://t.co/3dw9XaFyF3

[Peter Robinson @pnrobins]

Translating long COVID manifestations into computable HPO terms:
https://t.co/i5uwvh5Tyb @hp_ontology @MonarchInit