Taking single-cell RNA-seq by STORM: Van Andel Institute scientists answer your questions

It took a lot of time and effort. We performed a lot of hands-on testing with water first, then we incorporated different reagents from the kit, specifically with the varying viscosities. We had to confirm the delivery volumes were consistent and within the range we expected. Only then did we finally add samples into the protocol.

With the SPT Labtech mosquito HV, we have accomplished a six-fold reduction in the original volumes. Ideally, you should be able to do six-fold more reactions than the original kit.

We are currently limited by the number of unique indexes we can add to the plates for our pools. As soon as we have access to a larger number of indexes, we will work on creating larger pools of libraries and getting this in even faster throughput.

When looking at drop-out rates, there are a few things we need to take into consideration.

First, is the potential that the sorter may not successfully place a single cell in each well. From testing at our flow cytometry core facility, we expect less than 10 empty wells in each 384-well plate.

Another thing to consider is negative controls. In our 384-well plates, we include three negative controls that are leading to, in part, that 1% drop-out rate.

Lastly, there could be a potential issue with the genomics portion of the library prep. Altogether, we believe a 1% drop-out rate for the total workflow is very reasonable.

We use STORM-seq and 10x Genomics as complementary approaches for answering very different kinds of questions. We use 10x Genomics when interrogating thousands to tens of thousands of cell types to understand the cell-type diversity of our populations.

On the other hand, we use STORM-seq as a targeted, hypothesis-driven approach. You get thousands more genes with STORM-seq compared to 10x Genomics. That can be advantageous to capture some of the intermediate cell types that are otherwise missed with 10x Genomics, especially if these "marker genes" are driven by lincRNAs. A great example of this is the fallopian tube epithelium.

We are in the process of looking at repeat space because we have 2 x 100 reads and are probing the total transcriptome. We don't have data to share yet, but STORM-seq could prove useful in looking at this information.

Compartmap also works with single-cell RNA-seq, single-cell ATAC-seq, and whole genome bisulfite sequencing of single cells. It also works with methylation arrays. Beyond that, there are a variety of different single-cell profiling approaches that compartmap is compatible with to provide higher-order chromic confirmation.

In collaboration with Takara Bio and SPT Labtech, we are adding unique molecular identifiers (UMIs) into the mix, as well as some spike-in controls. What originally sparked the development of STORM-seq was the desire to profile the entire transcriptome combined with a multi-omic profile through orthogonal measurements using ATAC-seq or bisulfite sequencing of a single cell. We're working on this all-in-one method, named SCRAP-seq, and it will probably come out next year.