Johannes Köster

@johanneskoester

Algorithms for reproducible bioinformatics, Institute of Human Genetics, University of Duisburg-Essen

Essen. Germany
Vrijeme pridruživanja: siječanj 2016.

Tweetovi

Blokirali ste korisnika/cu @johanneskoester

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @johanneskoester

  1. 3. velj

    Recently, our daughter Flora Amélie was born. I will now be on parental leave for a month. Looking forward to merge your PRs in March ;-).

    Poništi
  2. proslijedio/la je Tweet
    26. sij
    Odgovor korisniku/ci

    After 10+ years of writing Python, I love that Rust is forcing me to think in a very different way. I also appreciate rust-bio, which I’ve been using for examples of idiomatic Rust related to my problem domain.

    Poništi
  3. proslijedio/la je Tweet
    25. sij

    This is what it looks like to push 750 Gbytes of wheat exome data through a 50 node, 2.4k CPU auto-scaling cluster on . Made easy with and a 6,670 job pipeline

    Poništi
  4. 21. sij
    Poništi
  5. proslijedio/la je Tweet
    20. sij
    Poništi
  6. 20. sij

    Here comes the brand new official 2min intro: . Thanks to !

    Poništi
  7. 20. sij
    Prikaži ovu nit
    Poništi
  8. 20. sij

    5.10.0 is released. It introduces integration, bringing the best of two worlds together: Scalable, modular, and readable data analysis definition with and fast interactive prototyping of individual steps (e.g. for plotting) with Jupyter. 1/n

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    14. sij

    Thanks to our expert for sharing his latest tool for variant calling! Great to close the scientific program of the WGGC2020 News&Views with also applications!

    Poništi
  10. 5. sij

    Heading to from in now. there needs to be a night train for such trips. It is crazy to take a flight and then sleep in a hotel for having a business in next morning. It can be more expensive, if it is as comfortable as a hotel.

    Poništi
  11. proslijedio/la je Tweet
    28. pro 2019.

    Want to join a funded project that studies multiple stressor effects on streams? Great project team 1 PhD pos. open - apply by Jan 10th

    Poništi
  12. 12. pro 2019.

    0.43.0 is released. This is the biggest release so far! It adds lots of tools contributed by the community! Docs: Release notes:

    Poništi
  13. proslijedio/la je Tweet
    7. pro 2019.

    Biologists should learn how to code: analyses should be performed with version controlled reproducible workflows (such as ' Snakemake). This requires rudimentary programming skill but it's much more efficient (for the author and users).

    Prikaži ovu nit
    Poništi
  14. 5. pro 2019.

    With such a definition, what you get here is for each variant a posterior probability for being present AND a posterior for being an FFPE artifact. Then, one can use the FDR control mechanism of to filter those FFPE artifacts away while considering uncertainties.

    Prikaži ovu nit
    Poništi
  15. 5. pro 2019.

    1.5.0 is released. The now allows expressions for selecting particular substitutions. This can e.g. be used to model FFPE artifacts in an uncertainty-aware manner:

    Prikaži ovu nit
    Poništi
  16. 4. pro 2019.

    I am proud to say that this also means that appears to be pretty close to the amazing in terms of popularity nowadays. Here are the stats for their 2018 paper: .

    Prikaži ovu nit
    Poništi
  17. 4. pro 2019.

    has just reached 500 citations! 341 of them in the last two years. This makes on average ~3 new citations (i.e. ~3 new driven published data analyses) per week:

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet
    29. stu 2019.

    "The previous PhD student used a self-developed code. I have managed to get the code(s), however, it is not working and gives multiple errors when I try." Students and scientists need training in coding etc to mitigate time sinks like the above.

    Poništi
  19. 28. stu 2019.

    1.4.0 now also supports MNVs. Moreover, it splits variant calling into a per-sample preprocessing and the actual joint calling, thereby improving scalability and allowing to quickly process multiple definitions.

    Poništi
  20. proslijedio/la je Tweet
    15. stu 2019.

    ⭐️ So cool to see a blockchain approach for caching and reusing intermediate results between workflows 🤓 The cache can be shared -- for everyone in your group, institute, or computing environment 🤯

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·