original preprint thread here x.com/russelltwalt...

It’s been a pleasure to work with the screening community at @broadinstitute.org and beyond to develop and apply CROPseq-multi over the past year, and we’re excited to continue to do so moving forward.

For combinatorial screens, a screen of previously reported synthetic lethal paralog pairs suggested a design of just six constructs per gene pair was sufficient to identify these genetic interactions.

For single-target screens, just one construct, the top a priori ranked guide pair from Broad GPP’s CRISPick, provides strong performance.

We observed robust performance for both guide positions, all three validated tRNAs, and with diverse 12 nt iBAR sequences.

Third, we share additional pooled viability screens evaluating vector performance for CRISPR-KO, CRISPRi, single-target, and combinatorial pooled screens.

Second, we validate a scRNA-seq readout based on 5’ direct-capture with high capture rates.

Our updated design, CROPseq-multi-v2 (Addgene: www.addgene.org/225752/), retains all the functionality of the original CROPseq-multi, while adding compatibility for methods based on T7-IVT-based barcode detection approaches, like PerturbView and NIS-Seq for optical pooled screens.

Our update has three major additions: (1) compatibility for T7 in vitro transcription (IVT)-based detection methods (2) a validated scRNA-seq workflow, and (3) additional pooled screens validating performance for CRISPR-KO, CRISPRi, and combinatorial screens.