Piotr Stępniak

In just 1 day for #AI peer-review publications (deep learning algorithms vs clinicians):
—3 new papers—@NatureMedicine X2, @AnnalsofIM (a 21% jump)
—1st prospective clinical validation (GI polyps)
—-broadening w/simulation of clinical urgent referrals (eye & neuro events)

The darn exosomes, that function like drones, blocking immunotherapy for #cancer. And perhaps a biomarker to guide responsiveness for anti-PD-1.
https://t.co/0B7A8Q9qG1 @Nature by @PennMedicine

The FDA approvals for #AI in medicine are accelerating.
@US_FDA @aidocmed @ZebraMedVision @baylabsinc @NeuralAnalytics @icometrix @Viz_AI @ArterysInc @maximumqai @AliveCor https://t.co/RbgPnIcOoo https://t.co/HX0Sd3WXH2
now ≥ 1/month; 10/13 scans, 1 eye disease, 1 neuro,1 heart

.@IBMWatsonHealth, we have a problem.
https://t.co/TrAa9fQoU9
"There a lot of promise for #AI—that promise is not realized"—John Kelly, head of @IBMResearch, a mea culpa
by @danielas_bot @tedgreenwald

Polish chemist and physicist Maria Sklodowska-Curie has made the biggest impact on world history according to new BBC poll - Women Who Changed the World: https://t.co/LslOzfZWv4

Redefining "deep phenotyping" and rebooting developmental biology w/ #CRISPR in vivo bar-coding every single cell @geochurch @harvardmed @wyssinstitute @sciencemagazine
https://t.co/pnUC3qDQHB

Using the liquid biopsy in #cancer to determine tumor mutational burden (TMB): utility and prognostic value for immunotherapy https://t.co/yBvnGRRqaX
@NatureMedicine by @drgandara @ucdavis @UCDavisHealth @genentech

—I'd like a map of the regulatory genome at the single-cell level
—No problem, here it is 😉
https://t.co/haaDcX2Z4z
@JShendure @coletrapnell @cusanovich @ahill_tweets and @UW team strikes again! @CellCellPress

Using single cell RNA-seq to discover and characterize an unknown cell type
https://t.co/VuPX0h5oqU @NatureNV
about these 2 @Nature new papers:
https://t.co/11K42AWrRA
https://t.co/ysYyPX8lOP

Thinking of the gazillion human gene-gene interactions we don't know about yet, admiring this #CRISPR platform work for going after over 200,000 gene pairs https://t.co/ug4nxuevis @CellPressNews @CellCellPress by @UCSFMedicine @UCSF @GladstoneInst @cziscience et al