Tue Jul 11, 2017 10:41 pm by AdeebR
Hi Mayan,
While I agree with Chad and Greg that you can certainly normalize for drifts in in signal based on the beads, it's important to appreciate that decreasing signal intensity does likely mean a lower signal to noise ratio, particularly for low intensity cell signals, and that normalization ultimately can't compensate for increased noise in the data.
As a result, when doing long barcoded acquistions we typically keep our bulk sample in PBS and prepare ~1-2M cell aliquots in H20, which we acquire in ~1hr increments.
After each acquisition, we run a QC check on the data and if the EQ bead signal intensity has dropped below a defined level (e.g., 20% of starting value, which usually takes ~4-5hrs of acquisition time), we pause acquisition and retune to reoptimize DV. Since we typically switch back to loops to retune, we simultaneously run a clean cycle on the SS.
Adeeb
Adeeb Rahman
Icahn School of Medicine at Mount Sinai, NYC