Jeremy Leipzig

@jermdemo

Bioinformatics engr & dev. PhD candidate & adjunct. O’Reilly author. Occasional crank. Pipelines, metadata, genomics, reproducible research.

Philadelphia, PA
Vrijeme pridruživanja: listopad 2008.

Tweetovi

Blokirali ste korisnika/cu @jermdemo

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

  1. Prikvačeni tweet
    25. tra 2019.

    "data available upon request"

    Poništi
  2. prije 22 sata

    Hey how involved was Cloudera in 2015 at the start of your project? I'm trying to piece together some ancient history.

    Poništi
  3. 31. sij

    I like how this entry ends with a lighthearted pumpkin anecdote after describing the espionage charges

    Poništi
  4. 30. sij

    Antoni was also involved in the first sequencing of HIV (LAV/HTLV-III)

    Poništi
  5. 23. sij

    anyone have a mnemonic to remember which is a forward and reverse genetic screen?

    Poništi
  6. 23. sij

    this PR says they did do a paired blood/stool assay with great results. the specificity/sensitivity numbers are not concordant with the poster but maybe they refer to training data?

    Prikaži ovu nit
    Poništi
  7. 22. sij

    The latest Freenome and Grail abstracts on CRC/GI cancers are up. Maybe they should also try the fecal route if their primary goal was to beat Cologuard?

    Prikaži ovu nit
    Poništi
  8. 22. sij

    They are taking blood from someone's arm, looking for tumor dna fragments, and trying to compete with a stool sample for detecting colorectal cancer. Very difficult. Just not enough signal.

    Poništi
  9. 21. sij

    Intriguing approach mapping papers to ontologies using NLP

    Poništi
  10. 20. sij

    This is a decent list I guess. I might have combined pipeline frameworks/cloud/docker/bioconda into one infrastructure item. The rise of centralized resources like ExAC and data commons. And then maybe Spark and exotic hardware?

    Poništi
  11. 16. sij
    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    15. sij

    Actually, is anyone in or using both {drake} and snakemake in a single project?

    Poništi
  13. 14. sij
    Prikaži ovu nit
    Poništi
  14. 14. sij

    I never understood why people faithfully put "Scientist II", "Engineer III", "Schmoogle I" on their LinkedIn profile like they really buy into some bullshit hierarchy imposed by their employer.

    Prikaži ovu nit
    Poništi
  15. proslijedio/la je Tweet
    13. sij
    Poništi
  16. proslijedio/la je Tweet

    I think there is a sample swap with my genome because 1) genotype concordance with 23andme data is too low 2) Mendelian inheritance error with my parents' genome is too high 3) PCA puts me near EUR in 1kgenomes although my parents are correctly in the EAS cluster

    Prikaži ovu nit
    Poništi
  17. 12. sij

    Do any journals accept tags, or at least not massacre them?

    Poništi
  18. 10. sij

    amazing. I bricked my wife's iphone trying to change the battery, but they can pull this off.

    Poništi
  19. 8. sij

    Is there a name for the type of white-glove advocated peer review that reproducible analyses enable, where the reviewers directly "kick the tires" of an analysis and help to improve it? envisioned this a few years back

    Poništi
  20. proslijedio/la je Tweet
    2. sij
    Odgovor korisnicima

    Code + data + (and in many cases) quite a lot of compute is needed for reproducibility. The first two are mentioned in the FAIR post, but who will reproduce all the huge projects?

    Poništi
  21. proslijedio/la je Tweet
    1. sij

    "The code used for training the models has a large number of dependencies on internal tooling, infrastructure and hardware, and its release is therefore not feasible."

    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:

    ·