The idea is to use rodents and to induce HGT within their microbiomes, one high exposure group, one low exposure and one control. The extent of HGT will then be assessed. What size would my sample have to be at a confidence interval of 95% to indicate a statistical significant difference/correlation. Seeing as my actual samples would be the stool samples collected and the quantified microbiome rather than the actual rodent itself. The rodents are in actual fact vessel that house the actual "study sample".