Skip to main content

Archived Comments for: Using PPI network autocorrelation in hierarchical multi-label classification trees for gene function prediction

Back to article

  1. Some remarks

    Fabio Fabris, University of Kent

    10 December 2013

    In the 16th page authors state that: "In this case [highly connected genes for both training and testing], there is a clear advantage of using NHMC with ¿ = 0 over using NHMC with ¿ = 0.5."

    However, it seems that the second part of table 2 (which should be split in two different tables imho) presents the opposite.

    Also, the values on the vertical axis of Figure 2 (varying between 0.2 and 0.4) are inconsistent with the values depicted on the tables.


    Thank you!

    Competing interests

    None declared

Advertisement