I'm performing single-nuclei RNA sequencing on different regions of the brain and spinal cord of post-mortem brain tissues. I already sequenced all my samples and my results have significant variability in terms of the number of nuclei captured, sequencing depth between samples (reads per sample), number of reads per nuclei, the median number of UMI per nuclei. My questions are as below:
1. Which sequencing parameters are more important to be consistent between samples that you want to integrate?
2. What should I do to compensate for this variability? for instance, if I can sequence some libraries for the second time which parameter should I consider to compare between samples, and then by sequencing more I could reach more consistency? I mean to help me make a better decision about which samples I need to sequence more.