Thinking about ways to finish the phrase "data scientists should ___ rather than becoming dilettante epidemiological model/curve fitters on case/mortality figures"

- write visualisations for emerging CoV-2 structural biology results
ー make interfaces to CoV knowledge graphs
ー Organise the literature by some experimental new SciBERT vector idea
ー Thread together literature feeds from across multiple journals
ー Put PDF screenshot snippets on social media based on salience measures

…basically, very quickly just becomes a wish list of things I want
Signal boosting this 🗣

This needs some coordination and focus if anyone is looking for something to work on… Ideally if you have background in bio but all it takes is to understand role of enzymes and catalogue individual mentions in emerging literature https://twitter.com/biochemistries/status/1248925038679556096
Yesss this is sounding more like it https://twitter.com/JedMSP/status/1249837950583451648
Here are 4 COVID-19 literature catalogues I've seen so far, just to keep this thread in one place: https://twitter.com/biochemistries/status/1250718850028625921

Notably the only "data science"-y part here is UMAP/t-SNE topic clustering — great work but I'd like to see more of an explorable @etymo_io-like solution 🤔
You can follow @permutans.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: