ImRNA or RNA antibody model? Donald Forsdyke, Queen's University, Canada 9 October 2009 The author's appendix (p. 11) summarizes nine steps in the proposed "imRNA response mechanism." Two of these are weak. Step 3: Recognition of foreign viral mRNA "possibly by host reverse transcriptases RT." But how does reverse transcriptase distinguish between self and not-self RNAs? Step 6: Integration into host genome. But there is no distinction between somatic cells and germline cells. How is the "Weismann barrier" breached in the case of germline cells? As Koonin points out, there is Lamarckian element here. While there is much indirect evidence, ably summarized by Flegel, that support his model (especially bacterial CRISPR systems), the above objections are overcome in the "RNA antibody" model we suggested (see Paper in Trends in Immunology . The model was updated in Chapter 12 of my textbook . 1. Forsdyke DR, Madill CA, Smith SD: Immunity as a function of the unicellular state. Trends Immunol 2002, 23:575-579. 2. Forsdyke DR: Evolutionary Bioinformatics 2006, Springer, New York. Competing interests No competing interest.