In high-throughput analysis, tabular data arise from counting reads (classified by gene), assigning samples to genotypes (often at many loci) or other types of markers.