meme in #rstats

"I tried loads of different analyses and chose the best one" Statistics: oooh that's really bad Machine learning: standard

The only way to achieve universal open access to the scientific literature is for research funders to require it of their grantees. They're finally doing it, but are taking a LOT of flak. Show your support by signing this open letter michaeleisen.org/petition/

A "worrying analysis": "18 [#deeplearning] algorithms ... presented at top-level research conferences ... Only 7 of them could be reproduced w/ reasonable effort ... 6 of them can often be outperformed w/ comparably simple heuristic methods." Paper: lnkd.in/dTaGCTv #AI

Nascent RNA analyses: tracking transcription and its regulation go.nature.com/31yoSTC #Review by Erin M. Wissink (@rnaErin), Anniina Vihervaara, Nathaniel D. Tippens & John T. Lis

Bien qu’elles mettent souvent en avant un numéro surtaxé, toutes les enseignes possèdent forcément un numéro non surtaxé que leurs clients peuvent utiliser dans certains cas.L'UFC - Que Choisir vous délivre la liste des principaux numéros gratuits pour ne plus payer inutilement.

A Single-Cell Transcriptome Atlas for Zebrafish Development dlvr.it/RBVNMv

50 years of eukaryotic transcription [collection of perspective and review articles by Robert Roeder, James Kadonaga, Ronald & Joan Conaway, John Lis] nature.com/collections/ga… #TFbinding #txnJC

Savez-vous ce que signifie le @ et depuis quand il existe ? En fait c'est la fusion du "a" et du "d" latin, donc du "ad" (vers, chez...). Le @ de votre adresse mail signifie bien "chez" tel ou tel domaine ! Le @ n'a rien d'une invention récente, il existe depuis le VIe siècle.

Je suppose que cette promesse n’a pas de portée rétroactive ? 🤔 🤷🏿‍♀️ #CNNum #voeuxMacron x.com/emmanuelmacron…

Monsieur Poulpe Un conseil que j'te donne mon bon Poulpe, quand tu seras au fond du trou, enfin si t'y arrives, ba pour remonter t'auras qu'à creuser dans l'autre sens ! (Méthode ancestrale niveau 12)

New research: Improving pairwise comparison of protein sequences with domain co-occurrence plos.io/2AG7b6M Menichelli et al.

New Research: Probing instructions for expression regulation in gene nucleotide compositions plos.io/2CEL9Di Bessière et al.

We've taken our greatest scientists and turned them into grant-writers. We've taken our greatest lawyers and turned them into merger and tax avoidance specialists. We've taken our greatest minds and pointed them at getting more clicks on ads.

Et voilà lvl 1 - 110 en finissant sur un lvl up 110 par une exploration ! Merci à ceux qui sont passés sur le stream et merci à Ancien compte de @ZeratoR d'être passé aussi :D Timer: 26:43:38 À battre pour les motivés :D

catg() { cat $1 | grep -v ">" | GREP_COLOR="01;36" egrep --color=always "T" | GREP_COLOR="01;31" egrep -i --color=always "A" | GREP_COLOR="01;37" egrep --color=always "G" | GREP_COLOR="01;32" egrep -i --color=always "C"; } catg seqs.aln

Lets "Put science first and formatting later" Our proposal for a format free submission policy with Aziz Khan and Alejandro Montenegro m.embor.embopress.org/content/early/…

Five ways to fix statistics: nature.com/articles/d4158… "Biomedical research generally uses the 2 sigma (P ≤ 0.05) rule; physics requires at least 3 sigma (P ≤  0.003)." As dataset are bigger and bigger, why biologists are not using the p<0.003 threshold, same as physicists?

#Preprint isn't a synonym for "first draft" or "low quality". In fact, we think scientists who post to a preprint server are taking #peerreview a step further by inviting MORE feedback from the community. journals.plos.org/ploscompbiol/a… #PowerToThePreprint