Using machine learning and computer vision techniques to automatically identify and quantify all intracellular substructures & organelles in isotropic EM data
Some new updates on the site, including rearranging and cleaning up the dataset pages πͺ.
Also, did you know that you can open (or view) any of the datasets directly in @FijiSc using either File/Import/N5 or Plugins/BigDataViewer/N5 viewer? Check it out!
Come work with us at @HHMIJanelia with the new π Annotation & Analytics π team tinyurl.com/5xxy2v95. This is a β¨ NEW β¨ multi-disciplinary team that will work with a variety of datasets (1/6)
#NewFeature alert! Now you can filter and sort available datasets on OpenOrganelle.janelia.org. Happy browsing! Thanks @neomorphic and @davisvbennett for the continuous improvements to the site π
We are looking for postdocs to do cool stuff with our upgraded mesoscope @HHMIJanelia. Now up to 50,000 neurons recorded simultaneously @ 3Hz in two planes. Full ad here: janelia.org/lab/pachitariu-lβ¦
2 Papers @Nature on the use of enhanced FIB-SEM to image cells and tissues at up to 4nm per voxel. The authors' AI-based analysis pipeline, which they used for reconstruction of 35 organelle classes, is available via OpenOrganelle
go.nature.com/3lhin4wgo.nature.com/3oEsthG
Itβs official! Our manuscript is out today in @Nature. π₯³ Additions since our preprint include a comparison of generalists trained on all training data vs specialists trained on subsets (spoiler alert: generalists win!) and #moreGolgi!
rdcu.be/cy1UJ (1/3)
We are proud to share our work on segmenting up to 35 organelles & macromolecular structures in whole cell isotropic FIB-SEM at 4 & 8nm resolution biorxiv.org/content/10.1101/β¦ 1/12
Weβve also gotten a facelift! πββοΈ Browse all the related data for these papers and more - AND - volume renderings are now available for most segmentations. Check it out on openorganelle.janelia.org !
Itβs official! Our manuscript is out today in @Nature. π₯³ Additions since our preprint include a comparison of generalists trained on all training data vs specialists trained on subsets (spoiler alert: generalists win!) and #moreGolgi!
rdcu.be/cy1UJ (1/3)
You might also consider using the free version of syGlass. syglass.io/view
It's significantly better with a VR headset, but the rendering is so choice. [Data credit to @COSEMdata ]
Looks like I'll be spending a lot of time playing around with openorganelle.janelia.org/ soon!
Great talk from @avweigel with such beautiful pictures (as always) at #AIM2021
what if you could explore biology in 3D, utilizing real data from microscopes? what if experts could easily create compelling VR experiences with their own data?
new version of @syGlassVR dropping in February π¬