You are conducting a large microbiome study and examine 2,000 bacterial species to test for association with a disease. If you were to test each species separately and at the a=0.05 level, how many of these markers do you expect to find evidence for association with disease (rejecting the null) by chance alone? In other words, on average how many type I errors would you most likely make, assuming none of these bacterial species are meaningfully associated with disease? a. 50 b. 5 C. 100 d. 250 e. 1000 f. 1900