In #statistics #jargon, I always struggle to remember which is which between "type I" and "type II" errors (a.k.a. "alpha level" and "beta level"), because these names are so arbitrary. But I never fail to grasp the difference between "false positive" and "false negative", for the obvious reason that these terms are self-explanatory.
Wikipedia has a few words explaining why there is a difference between "type I error" and "false positive" despite being mathematically equal: https://en.wikipedia.org/wiki/False_positive_rate#Difference_from_%22type_I_error_rate%22_and_other_close_terms
#Bioconductor 3.8 was released today! Within it, the update of the CAGEr package (to analyse activity of gene promoters) brings a few bug fixes and new syntactic possibilities in the plotAnnot() function, using formula to express combinations of metadata. https://bioconductor.org/packages/3.8/bioc/html/CAGEr.html
A new #RNA cap structure published by Munir, Banerjee and Shuman (Sloan-Kettering Institute, New York) in #NAR (https://academic.oup.com/nar/article/46/18/9617/5092737). All is in the title: « NAD+-dependent synthesis of a 5′-phospho-ADP-ribosylated RNA/DNA cap by RNA 2′-phosphotransferase Tpt1 »
The Open Material Transfer Agreement Show more
The Open #MTA was announced in this week's issue of Nature Biotechnology. Its goal is to promote and ease the exchange of research materials under the same spirit as #opensource software. This is one more step towards #openscience. See https://biobricks.org/open-material-transfer-agreement/ and https://www.openplant.org/openmta/
First use of the #CRISPR #Cas9 technique in #Oikopleura dioica, where double-strand breaks induce microdeletions when flanked by microhomology regions, because of the absence of the canonical non-homologous end-joining pathway. Deng, Henriet and Chourrout, 2018, Sars International Centre for Marine Molecular Biology, University of Bergen, Norway, https://linkinghub.elsevier.com/retrieve/pii/S0960-9822(18)31132-1
Journée Francophone de la Recherche à #Tôkyô le 12 décembre 2018. Inscriptions : 1er novembre 2018 (contributions) / 03 décembre 2018 (auditeurs).
The best French-speaking symposium in Japan! Joke apart, it leverages the modest but significant size of our linguistic community to make a truly #multidisciplinary event.
With all due respect for the contribution of #Latin to #science, I would like to stop using "et al." in #bibliography. (https://en.wikipedia.org/wiki/List_of_Latin_phrases_(E)#et_alii). This complicated way to say "and collaborators" is only a small obstacle for newcomers, but why keeping it in our digital age where a few extra characters will not cost precious ink and parchment anymore? I will probably go for "and coll." when not needing to follow an established style, unless there is a better alternative?
Xdrop: targeted sequencing of long DNA molecules from low input samples using droplet sorting. Show more
Article on #bioRxiv by Madsen, ..., and Mikkelsen (Samplix, Denmark). Enrichment by encapsulating #DNA fragments in double-emulsion #droplets, detecting the target by hydrolysis probe PCR, sorting the fluorescent droplets and amplifying their content by isothermal strand-displacement, and finally sequencing. When enriching for #HPV DNA in HeLa genomes, 5 to 10% of the reads were on target. https://doi.org/10.1101/409086
One step further towards entirely #synthethic #biology: Dou, collaborators and Baker published “De novo design of a fluorescence-activating β-barrel” in this week's issue of #Nature. Quoting their abstract: “This de novo design of small-molecule binding activity [...] should enable the design of increasingly sophisticated ligand-binding proteins, sensors and catalysts”. Meaning: computer-designing new enzymes instead of searching for them in natural diversity hotspots. https://doi.org/10.1038/s41586-018-0509-0
The authors of the #Nature article “Oxygen regulation of breathing through an olfactory receptor activated by lactate” report in another article in Nature that they can not reproduce their main finding (a defect in the hypoxic ventilatory response of mutant of that olfactory receptor), which invalidates the “Oxygen regulation of breathing through an olfactory receptor” claim that makes most of their manuscript's title, but still they did not go as far as its #retraction. How come ?
<Multi_key> <1> <1> <1> : "′" Show more
I have put this line in the .XCompose file of my home directory to type the "prime" character with the UNIX compose key. In molecular biology, we use a lot this character to indicate the direction on a #DNA strand, which is either « 5′ » or « 3′ » in our jargon. While a single quote character (« ' ») is often used as an ersatz, I prefer to enjoy nice #typography. (BTW, do not forget « include "%L" » on top of the file.)
Earth Biogenome: an ambitious project to #sequence the #genome of all living species on #Earth (actually, only eukaryotes; that excludes bacteria and archaebacteria, but includes unicellular organisms like yeast). Estimated data production in 10 years: 20 petabytes, well within the capacity of scientific infrastructures. https://www.earthbiogenome.org/
Distance to nuclear speckeles measured by quantitative sequencing shows stronger transcription deeper in nucleus. Show more
Mapping 3D #genome organization relative to nuclear compartments using TSA-Seq as a cytological ruler. http://jcb.rupress.org/content/early/2018/08/27/jcb.201807108
Chen, ..., and Belmont (Univ. Illinois) just published TSA-Seq (Tyramide Signal Amplification-Sequencing). Distances between chromosome domains and nuclear #speckles are reproducible across experiments; transcriptional activity increases from periphery to center.
Using Debian packages on CentOS with the Environment Modules system Show more
I am user on a computer #cluster running #CentOS, where we have to install software by ourselves. We use the "Environment Modules" system (http://modules.sourceforge.net) to manage parallel installs of different versions. Since I am a #Debian developer, I want to use our packages when possible. In the simplest cases I just have to extract the contents with "ar p packagename.deb data.tar.xz | tar xfJ -" and then write the "modulefile".
Researcher at OIST, Okinawa, Japan — Debian Developer. My private account is @charles_plessy on framapiaf.org.
This Mastodon instance is for people interested in technology. Discussions aren't limited to technology, because tech folks shouldn't be limited to technology either!
Hosting costs are largely covered by our generous supporters on Patreon – thanks for all the help!