Skip to main content

Archived Comments for: How accurate can genetic predictions be?

Back to article

  1. Brief comparison to Roberts et al.

    Jonathan Dreyfuss, Boston University

    6 August 2012

    <p>I hope you enjoy our article and welcome any questions/comments. <br/> <br/>One question I have received is how our paper relates to Roberts et al. [1]. Our article does not discuss their work, because we submitted approximately half a year before Roberts et al. came out. Briefly, Roberts et al. had a similar goal to ours: they tried to identify genetic prediction bounds without imposing a form on the risk distribution. However, Roberts et al.’s approach was found to be inferior to the liability-based analytical approximation, because Roberts et al. (a) do not account for identical twins’ shared environment (the liability-based approximation, and our method, use heritability to account for this), and (b) present "absolute upper bound[s]" that are lower than those from the liability-based approximation [2]. In turn, we directly compare the liability-based approximation to our method, and we show that the liability-based approximation's upper bounds are lower than ours. Thus, we feel that our paper does not suffer for lack of a comparison to Roberts et al. <br/> <br/>Sincerely, <br/>Jonathan M Dreyfuss <br/> <br/>References <br/>[1] N. J. Roberts, J. T. Vogelstein, G. Parmigiani, K. W. Kinzler, B. Vogelstein, V. E. Velculescu, The Predictive Capacity of Personal Genome Sequencing. Sci. Transl. Med. 4, 133ra58 (2012). <br/>[2] D. Golan, S. Rosset, Comment on “The Predictive Capacity of Personal Genome Sequencing”. Sci. Transl. Med. 4, 135le4 (2012).</p>

    Competing interests

    I am an author of this article.