Looking Review Paper About P-Value, Adjusted P-Value And Multiple Test In Bioinformatics

Question

[jack]

Hi everyone,

I'm looking for a review paper which gives comprehensive review about P-value, adjusted P-value and common statistical tests in bioinformatics.

It would be great if somebody give titles of good ones.

Answer 1

[brentp]

This may be somewhat tangential to your exact question, but you could start by looking at the series "Points of Significance" in Nature Methods (https://www.google.com/search?q=Points+of+significance).

They have a nice overview of some important ideas. E.g.: http://www.nature.com/nmeth/journal/v10/n11/full/nmeth.2698.html

There's not one explicitly about multiple-testing yet, but I suspect it won't be long.

[Emre]

Also see http://www.nature.com/news/scientific-method-statistical-errors-1.14700 for some insights on how to interpret P-values. One of the most striking remarks is the effect of sample size, that is, how you could get significant values with many data points comparing two distributions, although the difference between the means is tiny.

Answer 2

[Dan D]

It's not a journal article per se, but the book Intuitive Biostatistics by Harvey Motulsky provides an excellent overview of the concept of a P value in its first section, and then dedicates a section to the P value wherein he goes into about as much detail as you could possibly want.

Answer 3

[Andreas]

There is also 'How does multiple testing correction work?' (Noble, Nature Biotech, 2009)

Andreas

Answer 4

[Frédéric Mahé]

What about p-value misuse? A Dirty Dozen: Twelve P-Value Misconceptions (2008) Steven Goodman, Seminars in Hematology