Skip to main content

Archived Comments for: Subfunctionalization reduces the fitness cost of gene duplication in humans by buffering dosage imbalances

Back to article

  1. Diverse buffers for dosage imbalances

    Ariel Fernandez, The University of Chicago

    17 May 2012

    Expression divergence due to subfunctionalization, the focus of this contribution, is one of various regulatory mechanisms to buffer the effects of dosage imbalance resulting from gene duplication. There are other processes operating at the post-transcriptional regulatory level (protein oligomerization, micro-RNA suppressive control, post-translational modifications, etc.) that may serve as alternative buffers to mitigate the effects of dosage imbalance (A. Fernández and J. Chen, Genome Research 19, 2185-92, 2009). For this reason, we postulate two broad but distinctive distributions corresponding to two qualitatively different buffering regimes that are apparent in the plots of Fig. 1: One regime, in which genes exclusively or nearly exclusively resort to expression divergence as a buffer (datapoints tightly correlated along trend lines) and another regime where alternative post-transcriptional regulatory mechanisms participate in the buffering process to a significant extent in conjunction or to the exclusion of expression divergence (spread datapoints). As reliable quantitative data on alternative buffers becomes available, further studies from my laboratory will be forthcoming to address these issues and classify genes according to their buffering (single or mixed) strategies.

    There is a tight correlation between expression divergence within a protein family and our measure of dosage sensitivity for gene duplicates that rely on transcriptional expression divergence as the buffer mechanism. This tight correlation can be accounted for by noticing that cross-hybridization-related noise affects paralog genes to the same extent, and therefore, the noise can be filtered out if we restrict the analysis to transcriptomal paralog comparison. The computation of expression divergence through pairwise paralog comparison essentially cancels out the noise.

    The noise in the signals for two paralog genes 1 and 2 in a specific cell type can be decomposed into a background noise resulting from probe mishybridization with nonparalog genes and a noise contribution arising from cross-hybridization of the probe with all paralogs of the genes considered. Since the latter is proportional to the sum of noiseless signals for such paralogs, the expression distance between paralogs 1 and 2 becomes proportional to the difference in the noiseless signals for the two paralogs. That is, the noise is essentially filtered out as the microarray analysis is restricted to a paralog comparison within the transcriptome.

    Competing interests

    The author has no competing interests.

Advertisement