If you’re interested in trying MULTI-ATAC yourself, have a look at the protocol in the SI, and reach out to me for further details about oligos and barcode sequences.

MULTI-ATAC has already been successfully applied in other contexts as well. In this work, my collaborator Katya used MULTI-ATAC to probe the mechanisms by which deleterious mutations accumulate in the mitochondrial genome during aging:
www.biorxiv.org/content/10.1...

By including multiple replicates and doses per drug, we discover high-confidence drug-responsive chromatin regions and genes. We show that EZH2 degradation via MS177 coincides with up-regulation of NF-kB signaling, and SWI/SNF perturbation elicits a potent type I interferon response.

MULTI-ATAC is also compatible with the 10x Genomics Multiome kit. To demonstrate the type of high-throughput experiment enabled by MULTI-ATAC, we perform a 96-plex multiomic drug assay comparing inhibition and degradation of different epigenetic remodelers in a model of primary immune activation.

To combat this, we re-designed our MULTI-seq method and optimized it to enable pooled transposition of barcoded nuclei on the 10x Genomics scATAC-seq platform. We show that pooled transposition with MULTI-ATAC completely negates the batch effects that arise from processing samples separately.

The effect of variable transposition is well-established in bulk ATAC data but surprisingly has not to our knowledge been examined in scATAC-seq. This is especially important to consider when unique samples (i.e., different patients, disease states, etc.) are transposed in a parallel format.

For example, we see that variability in Tn5:nuclei ratios unsurprisingly skews per-nucleus fragment yields. However, this impacts dimensionality reduction and - importantly - even biases cell type recovery.

scATAC-seq data across different technologies and sample types suffer from batch effects that are linked to transposition reaction conditions and directly impact biological interpretation.