The Estimation of Concentration of Viruses and Bacteria from Dilution Counts

Abstract

The method of cal-culating the estimate of the density of a suspension of infective particles from parallel counts for a series of independent dilutions of the original suspension is investigated. It is shown that for this estimate of density the method of maximum likelihood has a very simple solution, viz. the sum of the counts over all plates and dilutions divided by the sum of their corresponding dilutions. This estimate is readily calculated and has greater efficiency, or smaller variance, than the commonly used arithmetic mean of the densities obtained for individual dilutions. The underestimation of density arising from inclusion of counts affected by overcrowding and/or clumping at high concentration is eliminated by a test of linearity of mean count on dilution. This is done by constructing successive orthogonal comparisons to test the counts in a stepwise fashion from the least to the greatest concentration to check whether any concentration is so great as to lead to underestimation of density.