Favorite Tweets from Late Nov. '17 Meetings - ERCC9, RSG '17+ misc (#ERCC9,i0exrna17,#RSGDream17, i0rsg17)

@FertigLab Changes in expression come before changes in methylation, contrary to their initial expectations. #rsgdream17

.@FertigLab cites PatternMarkers & GWCoGAPS for novel data-driven biomarkers via whole transcriptome NMF  https://academic.OUP.com/bioinformatics/article/33/12/1892/2975325 … Sparsely decomposes gene expression into patterns #rsgdream17

.@FertigLab presents dynamics of therapeutic response in cancer. Focusing on EGFR inhibitors #rsgdream17

Ziga Avsec presents his poster "Modeling positional effects of regulatory sequences with spline transformations increases prediction accuracy of DNN" at #RSGDream17 - paper now online  https://doi.org/10.1093/bioinformatics/btx727 …

Outstanding keynote by Christina Leslie on epigenomic signatures distinguishing 'reprogrammable' vs. 'non-reprogrammable' immune cells, and cell surface markers that can be used to distinguish them to guide cancer immunotherapy #RSGDream17  https://twitter.com/markgerstein/status/932329278729015307 …

@anshulkundaje @RalucaGordan I might have transcribed it wrongly but my impression was @RalucaGordan indicated that binding events in open chromatin regions appear "stronger" than otherwise #rsgdream17

Leslie cites "Chromatin states define tumour-specific T cell dysfunction and reprogramming"  https://www.Nature.com/articles/nature22367 … Has changing ATAC-seq peaks & gene expression betw. tumor & normal #rsgdream17

@manoliskellis @markgerstein You forgot F2F :)

Parida's talk has lots of useful #DataScience buzzwords: 3Vs for data (= velocity, volume, variety) + WfG (@IBMWatson for Genomics) as a transparent, comprehensive, navigable, & scaleable system + the hologenome for microbiome #rsgdream17

@IBMWatson Genomics for Cancer cure. Excellent insights by Laxmi Parida at #RSGDream17

Ooh. Interest piqued: Laxmi Parida talking about IBMResearch's findings working with the @Genographic Project genomics data ... #RSGDREAM17

@markgerstein After P2P and B2B, this is M2M :-)

Amazing talk on ribothrypsis by Manolis Maragkakis uncovers new mechanisms of ribo-stalling and mRNA degradation  http://mourelatos.med.upenn.edu/index.php/team/manolis-maragkakis/ … #RSGDream17

.@RalucaGordan reviews many key biases in chipseq - eg DS breaks depend on nt seq., PCR amplification, stronger signal related to open chromatin & high expression, different results w/ diff. antibodies (particularly variable in ENCODE comparisons for ATF2) #rsgdream17

Raluca Gordon: Chip-seq =/= in vivo binding. Combine multiple assays! #RSGDream17

Bayesian nets, linear deconvolution, adaptive multitask lasso & deep learning all before 12pm. Feeling at home at #RSGDREAM17 🤓

Nice keynote by @dana_peer at #RSGDREAM17 Single cell RNAseq expts w/ high dropout use a little MAGIC, good application of manifold learning

Chip-seq biases @RalucaGordan as comparing to cut&run from  https://elifesciences.org/articles/21856  #rsgdream17

#rsgdream17 made it to the fourth talk before a deep learning appearance (LINCS signature learning). I would have guessed fewer.

Castro in @RichBonneauNYU lab: Use TF activities to model gene expression via inferred networks. Using Adaptive Lasso (w/ a variety of sparse & block penalties) #rsgdream17

Single va Multitask learning and Network Inference from @RichBonneauNYU #RsgDream17

A MAGICal keynote. Now Ryan Peckner on deconvolution of next gen mass spec data #RSGDream17

Dana Pe'er: a potential downside of MAGIC is reduction of true stochasticity in the data #RSGDream17

From Peer, some key aspects of MAGIC: recovers gene-gene relationships, archetypical cell states & shape (manifold) data. Allows one see entire EMT from a single time point #rsgdream17

And here's the MAGIC pre-print  https://www.biorxiv.org/content/biorxiv/early/2017/02/25/111591.full.pdf … (Markov Affinity-based Graph Imputation of Cells) #RSGDream17

Great talks at ISG 5yr Anniversary! Characterizing genomes w/ new methods, software, sequencing tools @markgerstein @graveley @ENCODE_NIH

Posting talk tomorrow at 5-year celebration ceremony @UConn Institute for Systems Genomics  http://lectures.gersteinlab.org/summary/Genomics-n-DataSci--20171108-i0isg/ … #Genomics & DataScience

Janos Zempleni: Exosomes in cow milk and encapsulated RNA can survive pasteurization and are taken up by human (&🐁) intestinal cells! #ERCC9

Rozowsky: the pipeline ( http://github.gersteinlab.org/exceRpt ) hows differences in exo-/endo- genous profiles across body fluids #ERCC9

Looking forward to trying out exceRpt pipeline when back home... #ERCC9

Milosavljevic: Does the exRNA atlas lend itself to deconvolution? A: Carrier-specific exRNA profiles can detected across body fluids #ERCC9

Milosavljevic: ~100 miRNAs present in reasonably high abundance in each body fluid (except for urine) #ERCC9

attending #ERCC9. Looking forward to learning and sharing our capabilities on RNA and exosome detection at single particle level

Seda Kilinc Avsarglu: Overexpressed oncogenes are released from tumour cells in extracellular vesicles. #ERCC9

Dixon: shows that the RBP HuR enhances uptake of RNAs into EVs & is up-regulated in colon cancer #ERCC9

Daniel Dixon @KUMedCenter RNA binding protein HuR regulates EV secretion in colorectal cancer #ERCC9

Sukhbir Kaur @NIH_NCI: EVs sorted based on surface tags (CD63,CD47,MHC Class 1) have distinct ncRNA / miRNA profiles #ERCC9