Low no. of multiplets detected than expected

[AnjaliC4]

Hi,

I have noticed that no. of multiplets detected by AMULET are proportional to number of cells in singlecell.csv. For clarification, does amulet considers only the cells in singlecell.csv for testing multiplets or does it consider all the barcodes in the singlecell.csv?
Since I am not using 10x cell calling method, I would like for AMULET to detect multiplets from all the barcodes in the singlecell.csv. Is that feasible?

Thanks

[alperoglu]

Hi Anjali,

AMULET only considers the barcodes that have been labeled as cells with the is_cell_barcode column. You can include other barcodes in the singlecell.csv file by setting their values in the is_cell_barcode column from 0 to 1. However, by doing this you can be including empty droplets' barcodes which will have many reads that will not pass AMULET's quality thresholds. These empty droplets can negatively impact the performance of the algorithm. You can initially run the overlap counting portion of AMULET to get the summary statistics for your barcodes, reads and overlaps. These can give you more insight about which barcodes to consider or not.

Best,
Alper