RT @TrainingVbc: Congratulations to @FilipNemcko PhD (@stark_lab, IMP), who recently defended his thesis: 'Systematic identification and fu….

RT @deAlmeida_BP: Nature biotechnology highlighting ours @stark_lab and @steinaerts @ibrahimihsan recent work on designing cell type-specif….

RT @NatureBiotech: Research Highlight: Designer enhancers for cell-type-specific gene regulation

Cracking the 2nd code of life has been a major goal of @AlexanderStark8 and the @stark_lab since their start at the @IMPvienna @ViennaBioCenter in 2008. Big thanks to all members and collaborators, especially @deAlmeida_BP & @Eileen_Furlong who made this possible! (fin).

See also the related and complementary work by @ibrahimihsan @steinaerts (@VIBlifesciences) in fly brains – together, these publications are the first to read & write enhancer DNA in vivo. (9/N)

Our work demonstrates the feasibility of targeted enhancers design! Transfer learning using genome-wide and small-scale functional datasets should enable the design of tissue-, cell type-, and cell state-specific enhancers in any system. For more: (8/N).

We then design 8 synthetic enhancers for 5 tissues de novo (CNS, epidermis, gut, muscle and brain, 40 total) and test them in fly embryos in vivo. Of the 40, 31 (78%) are active and 27 (68%) function in the target tissue (100% for CNS and muscle) – we “wrote” enhancer DNA! (7/N)

@deAlmeida_BP first trains DNA accessibility models on genome-wide scATAC data and then adjusts these models by transfer learning to predict tissue-specific enhancer activity. The final models achieve high PPVs in CV, for example >70% for CNS and epidermis (6/N)

Today, we - @stark_lab and @Eileen_Furlong lab (@EMBLheidelberg) - report the design of 40 synthetic enhancers for 5 tissue types in fruit fly embryos, in vivo. With only a few hundred enhancers per tissue for training, we adopt a stepping-stone transfer-learning approach: (5/N)

Only with the availability of #DeepLearning #AI and #STARRseq, @deAlmeida_BP in 2022 managed to predict and design synthetic enhancers de novo for a single cell type (and #MPRA-derived synthetic enhancers can work in vivo, Gosai et al bioRxiv 2023) (4/N)

Scientists however cannot read & write enhancer DNA, nor evaluate mutations/SNPs in enhancers. A decade ago, we tested >7700 enhancer candidates in fly embryos (@BarryJDickson @evgenykvon) revealing rich sequence-to-activity maps – but still no code (3/N)

Our genomes contain about 20,000 protein-coding genes and probably millions of gene-regulatory regions called enhancers. In the 4-letter alphabet of the DNA, these regions encode tissue-specific gene expression patterns (see the 4 mouse enhancers from the VISTA Browser) (2/N)

We are very excited to present a major breakthrough achievement – the de novo design of synthetic enhancers for selected tissues in fruit fly embryos in vivo using deep- and transfer learning, @deAlmeida_BPet al published today in @Nature Thread 👇(1/N)

RT @biorxivpreprint: Developmental and housekeeping transcriptional programs display distinct modes of enhancer-enhancer cooperativity in D….

RT @jacobsjelle1: Excited to share the final version of my Postdoc work in the @stark_lab on mapping and unravelling the regulatory specifi….

RT @embojournal: Transcription pre-initiation complex formation & composition is not as uniform as we thought:.@stark_lab @IMPvienna finds….

RT @IMPvienna: 🚨 Research paper alert!.Functionally distinct #promoter classes initiate transcription via different mechanisms, according t….

RT @IMPvienna: Research paper alert! The @stark_lab shows in @genomeresearch that the activity of #enhancers is encoded through a combinati….

RT @TrainingVbc: Monika Heinzl from @jkulinz joined the lab of Alex Stark @stark_lab at @IMPvienna to understand and predict the DNA-sequen….

RT @TrainingVbc: The @stark_lab investigates how transcription is regulated at the enhancer and core-promoter DNA level & the proteins invo….