RT @sveta_kutuzova: (1/4) Our new preprint: a metagenome binner TaxVAMB that uses taxonomic labels to create better bins. In a nutshell, Ta…

@apcamargo_ @NatureBiotech @shaman_ns @DTU_HealthTech @NNFCPR @uni_copenhagen @DTU_Compute @ClinMicrobiom @foodDTUdk @UCPH_health @simonrasmu @JoJohanse Thank you! And thank you for your contribution to the software :)

@larsjuhljensen @krassowski_m @python_tip For Python:

A few months back, I wrote a blogpost on what you need to know about how computers work in order to micro-optimize code. Might be useful for other bioinformaticians.

@MandelKen @lh3lh3 @bioinfo_mark I think you might be seeing the result of which was recently pushed. Still CodecZlib uses a slow zlib impl. I would expect less than 10% diff b/w C and Julia on compressed data

@cgarciae88 Hi! I think you might have mistimed this. Jax computes the result lazily, so the function returns before the result is finished. On my computer, Jax takes 380 ms. Vanilla Julia takes 2000 ms, and with Julia packages, you get get it done in 300 ms :)

@noah_legall @AlbertVilella Yes. It manages to easy to learn, intuitive and concise, making it almost as productive as Python, while still being nearly as fast as C. It has the potential to be an end-all be-all language for bioinformatics.

@noah_legall @AlbertVilella We constantly face the two-language problem. We need an easy-to-write, expressive language because we do interactive data exploration, and to implement domain-specific algorithms outselves, but we also have large datasets and so need speed. Current C+Python doesn't cut it.

@jessenleon Hvis ikke det var for den røde linje ville jeg ikke kunne gætte hvilken vej korrelationen går... Mere retvisende at sige "Kvaliteten af undervisning har næsten ingen effekt på evalueringerne"? Stadig et ærgerligt resultat, selvfølgelig.

@polytomous Type 1+1 in a fresh console and it lags for half a second (slow, awkward start). Even worse for large libraries. Yet when it your code is up and running, it's lightning fast. "Luxury" because it's fun to write :)

Amazing speech by @Alfons_Valencia at @eccb18: One of the most basic things we teach students about transcriptomics is wrong. Most transcript isoforms don't matter.