Is it normal that nearly all of p-value is equal to 0.1 in PERMANOVA analysis
0
0
Entering edit mode
3.9 years ago
zhangdengwei ▴ 210

Hi all,

I'm computing the bacterial community significance among different groups using PERMANOVA analysis, but I got most of the Pr is equal to 0.1, and I got the following warning when computing it with R

'nperm' >= set of all permutations: complete enumeration.
Set of permutations < 'minperm'. Generating entire set.


And part of result is

group   distance    Df  Sums of squares Mean squares    F.Model Variation (R2)  Pr (>F)
EHCC-E/EHCC-I   Bray-Curtis 1   0.649072950650164   0.649072950650164   9.6075686839111 0.762047697282951   0.1
EHCC-E/EHCC-S   Bray-Curtis 1   1.03222905382541    1.03222905382541    14.2904148017771    0.781306217308349   0.1
EHCC-E/JC-E Bray-Curtis 1   1.01125210056938    1.01125210056938    17.5259268140954    0.814177571328508   0.1
EHCC-E/JC-I Bray-Curtis 1   0.931942533519936   0.931942533519936   19.337789036189 0.828604158097524   0.1
EHCC-E/JC-S Bray-Curtis 1   1.18680751735964    1.18680751735964    22.9827639530517    0.851757217794302   0.1
EHCC-E/JNU-E    Bray-Curtis 1   1.02925311435617    1.02925311435617    15.7396495394618    0.797362157215429   0.1
EHCC-E/JNU-I    Bray-Curtis 1   1.01206481882558    1.01206481882558    21.7998466018218    0.844960318495942   0.1
EHCC-E/JNU-S    Bray-Curtis 1   1.10600715442071    1.10600715442071    20.0110809048334    0.833410248549252   0.1
EHCC-E/MNIT-E   Bray-Curtis 1   0.841949076508153   0.841949076508153   15.7788623733998    0.797763899435413   0.1
EHCC-E/MNIT-I   Bray-Curtis 1   0.660537706742608   0.660537706742608   10.5753059567158    0.779010505577896   0.1
EHCC-E/MNIT-S   Bray-Curtis 1   0.948368532469721   0.948368532469721   16.4694284133139    0.804586629424479   0.1
EHCC-I/EHCC-S   Bray-Curtis 1   0.381554028042994   0.381554028042994   7.3439278043265 0.709974773920483   0.1
EHCC-I/JC-E Bray-Curtis 1   0.724004204343361   0.724004204343361   22.2229341534699    0.881060625946752   0.1
EHCC-I/JC-I Bray-Curtis 1   0.53427273741601    0.53427273741601    26.8445901874865    0.899479269738545   0.1
EHCC-I/JC-S Bray-Curtis 1   0.806241472500914   0.806241472500914   32.9113356193161    0.916460918307189   0.1
EHCC-I/JNU-E    Bray-Curtis 1   0.813741813448363   0.813741813448363   18.9970426548544    0.86361803052021    0.1
EHCC-I/JNU-I    Bray-Curtis 1   0.678889848888971   0.678889848888971   38.6925064808861    0.928044623525445   0.1
EHCC-I/JNU-S    Bray-Curtis 1   0.713699325644403   0.713699325644403   24.3265172486017    0.890216525848953   0.1
EHCC-I/MNIT-E   Bray-Curtis 1   0.354929347647845   0.354929347647845   13.2480409283906    0.815362355792813   0.1


The key code is

for (i in 1:(length(group_name) - 1)) {
for (j in (i + 1):length(group_name)) {
group_ij <- subset(group, site %in% c(group_name[i], group_name[j]))
otu_ij <- otu[group_ij$names, ] adonis_result_otu_ij <- adonis(otu_ij~site, group_ij, permutations = 999, distance = 'bray') adonis_result_two <- rbind(adonis_result_two, c(paste(group_name[i], group_name[j], sep = '/'), 'Bray-Curtis', unlist(data.frame(adonis_result_otu_ij$aov.tab, check.names = FALSE)[1, ])))
}
}


So, have I got the wrong Pr`? Thanks in advance!

PERMANOVA statistics metagenomics • 10k views
0
Entering edit mode

hi,whether you have solved this problem? i have the same problem with you!Could you tell me how to solve this problem? thank you !