🚨Preprint alert! tinyurl.com/2zkmynsd Did you ever dream of reconstruction #spatial single-cell #enhancer activity in a multicellular organism using massively parallel enhancer-reporter assays? Then Spatial-scERA is the method for you! #spatialOMICs

Very proud of this huge project led by Baptiste Alberti & Séverine Vincent, together with Isabelle Stevant! Great collaboration with @paulvilloutreix.bsky.social on the computational aspects.

We combined the principle of MPRAs with single-cell transcriptomics and spatial reconstruction using optimal transport to reconstruct virtual maps of enhancer activity in stage 6 Drosophila embryos. Our method presents 2 key innovations:

1) it combines scRNAseq with targeted PCR to improve the identification of cells in which an enhancer is active. 2) it uses a custom version of novoSpaRc, a computational method for spatialization to reconstruct a map of enhancer activity on a virtual embryo.

Thanks to these innovations, we could greatly improve cell recovery and faithfully reconstruct the spatial activity of enhancers by plotting their probability of presence at each position, even when very few positive cells have been sequenced.

We applied the method to 25 candidate enhancer regions in stage 6 Drosophila embryos and demonstrated that single-cell enhancer-reporter assays alone are not always sufficient to predict enhancer activity in vivo. In fact, specialization is really essential!

Just an example to showcase the power of our method: we could reconstruct the activity pattern of the h_stripe1 enhancer based on the sequencing of our reporter construct in only 9 cells!

Interestingly, our results highlight that chromatin accessibility is often a poor predictor of enhancer activity. Some of the regions we tested overlapped with highly accessible regions. However these regions are not active in reporter assays!