This dataset consists of raw nanopore signal data from biological replicates exposed to UV and antibiotic treatments, generated for microbial viability inference using deep learning models, as described in our manuscript Nanopore- and AI-empowered microbial viability inference (doi: 10.1101/2024.06.10.598221). It comprises one biological replicate (BR) of UV-treated Escherichia coli (dead, extracellular DNA from supernatant) and one BR of untreated E. coli (alive), as well as two BRs of UV-treated Chlamydia suis (dead) and two BRs of untreated C. suis (alive). We also provide two BRs of antibiotic-treated E. coli (dead, DNA from supernatant) and two BRs of non-antibiotic-treated E. coli (alive, DNA from pellet). The dataset includes two BRs of a quasi-mock community, each containing antibiotic-treated susceptible E. coli and antibiotic-treated resistant Klebsiella oxytoca.