bioinformatics logo

CGGC: Compare Groups of Growth Curves

Introduction

CGGC performs permutation tests to assess differences between groups of growth curves. It calculates all pairwise comparisons between two or more groups of growth curves.

To conduct CGGC on your own data, simply type data into the box below or paste data into the box from a text document or Excel spreadsheet.

Accurate p-values can be obtained by setting the number of permutations to some large value. We suggest 100 permutations for initial testing but at least 10000 permutations for definitive and publishable results.

Data Requirements:

The first row contains columns headers.
Each other row of data contains:
In the first column a group indicator (for example, mouse strain in the example below).
All other columns contain data for individual time points.
Missing data needs to be represented by the 2 letters "NA", without the quotes.

Some Example data is:
Strain     D0      D1      D3      D5      D7      D9      D11
BALB/c     4.01    3.96    2.62    2.59    2.13    1.47    0
BALB/c     4       3.91    3.94    3.81    2.82    1.74    0.98
BALB/c     4.32    4.93    4.15    NA      3.88    1.82    0.93
C57B/6     4.7     3.61    3.61    3.68    3.5     1.43    0
C57B/6     5.33    5.58    4.26    4.55    3.77    2.19    0.95
C57B/6     4.1     4.38    3.42    3.43    2.59    NA      NA
C.lmr1/2   4.57    3.51    3.72    2.55    2.05    0       0
C.lmr1/2   3.91    3.58    2.94    2.87    2.26    1.2     0.97
C.lmr1/2   4.79    3.82    4.21    3.78    2.43    1.75    0
The output from this example is:
 The number of lines of data processed (excluding the heading line) = 9
 The number of unique groups entered = 3
 The number of time points entered   = 7
 The number of permutations selected = 10000
       Group1    Group2    Stat  P.Value   adj.P.Value
  1    BALB/c  C.lmr1/2   0.603     0.52             1
  2    BALB/c    C57B/6  -0.598     0.64             1
  3  C.lmr1/2    C57B/6   1.203     0.39             1
(Your P values may vary slightly.)
Enter your data here:


Column data are separated by a tab , space , comma , or any whitespace .
Whitespace implies any number of spaces and/or tabs are allowed between columns.

  Enter the number of permutations - suggest 100 initially, 10000 for definitive results.

                             Display Data Values:

Authors

Method designed by Gordon Smyth and Russell Thompson with web interface by Keith Satterley.

Method

CGGC computes a permutation p-value for each pair of groups, using the average t-statistic between the groups as the test statistic. t-tests are computed for each time and averaged to obtain the permutation statistic. The actual computations are performed by the compareGrowthCurves function of the statmod R package.

The method is described and used in the references Elso et al (2004) and Baldwin et al (2007) listed below.

Citation

If you use results from this page in a publication, please describe the test as the "CGGC permutation test" and cite one of the following two papers:

  • Elso, C. M., Roberts, L. J., Smyth, G. K., Thomson, R. J., Baldwin, T. M., Foote, S. J., and Handman, E. (2004). Leishmaniasis host response loci (lmr13) modify disease severity through a Th1/Th2-independent pathway. Genes and Immunity 5, 93-100. [PubMed]
  • Baldwin, T., Sakthianandeswaren, A., Curtis, J., Kumar, B., Smyth, G. K., Foote, S., and Handman, E. (2007). Wound healing response is a major contributor to the severity of cutaneous leishmaniasis in the ear model of infection. Parasite Immunology 29, 501-513. [PubMed]

Comments/Questions? Contact Gordon Smyth.
Last modified: 17 March 2019