Many institutions will not let unpublished data leave the building — but you can read five numbers off the Cell Ranger summary. Everything below is computed in your browser: nothing is uploaded, stored or logged. If your data can be shared, the full report needs an upload.
Mitochondrial content is tissue-dependent: 5% is a sensible ceiling for blood, but applying it to heart or liver deletes the cells you are studying. The expected doublet rate follows how many cells you loaded — roughly 0.8% per 1,000 cells recovered on droplet platforms, so 10,000 cells means about 8%, and a gene-count ceiling will not remove them. Saturation tells you whether more sequencing would still buy you more genes. The long version is in how to set QC thresholds and defend them.