It is generally recommended to sequence 50 million or more reads/library-molecules per ATAC-seq sample for open chromatin detection and differential analysis (Buenrostro et al. 2015) and 200 million reads for TF footprinting (Yan et al. 2020). Preferred are paired-end sequencing data but single-end data are also usable. Most frequently ATAC-seq libraries are sequenced on the NextSeq with PE 75bp reads.
Paired-end data have slightly higher unique alignment rates, allow for PCR-duplicate removal and provide more complete information about the accessible sequences (due to the longer sequence information). Paired-end data allow for example the assignment of reads to categories such as nucleosome-free, mono-nucleosomal, and di-nucleosomal origins (Buenrostro et al. 2015) .
Controls are typically not run for ATAC-seq studies, but typically two biological replicates are required.