Question

我试图在素食主义者中使用嵌套的PERMANOVA，但是在F统计量和主要影响的p值方面，我得到的结果与biobiR和PRIMER-7的结果不同。任何帮助将不胜感激。

素食和生物多样性保护的结果：

library(vegan)
library(BiodiversityR)

# Make metadata
metadata <- cbind.data.frame(SampleID = seq(from = 1, to = 18, by = 1),
                             Diet = rep(c("Ref", "Soy"), each = 9),
                             Tank = rep(c("T1", "T2", "T3", "T4", "T5", "T6"), each = 3))

# Make abundance table
set.seed(1910)
low <- data.frame(replicate(100,sample(0:10, 9, rep = TRUE)))
colnames(low) <- c(1:100)

set.seed(1910)
high1 <- data.frame(replicate(90,sample(0:10, 9, rep = TRUE)))

set.seed(1910)
high2 <- data.frame(replicate(10,sample(10:30, 9, rep = TRUE)))

high <- cbind(high1, high2)

colnames(high) <- c(1:100)

table <- rbind(low, high)

# Calculate distance metrix
dist_bray <- vegdist(table, method = "bray")

### Nested PERMANOVA in vegan ###
# Define permutation scheme 
perm <- how(complete = TRUE, 
            within   = Within(type = "none"),
            plots    = with(metadata, Plots(strata = Tank, type = "free")))    

# Nested PERMANOVA 
adonis2(dist_bray ~ Diet + Tank, data = metadata, permutations = perm)
#> Set of permutations < 'minperm'. Generating entire set.
#> Permutation test for adonis under reduced model
#> Terms added sequentially (first to last)
#> Plots: Tank, plot permutation: free
#> Permutation: none
#> Number of permutations: 719
#> 
#> adonis2(formula = dist_bray ~ Diet + Tank, data = metadata, permutations = perm)
#>          Df SumOfSqs      R2      F Pr(>F)
#> Diet      1  0.22595 0.20758 4.1298    0.1
#> Tank      4  0.20602 0.18927 0.9414    1.0
#> Residual 12  0.65653 0.60315              
#> Total    17  1.08850 1.00000

### Nested PERMANOVA in biodiversityR ###
nested.npmanova(dist_bray ~ Diet + Tank, data = metadata, permutations = 999)
#> Total sum of squares of distance matrix: 1.088504
#> Total sum of squares for non-parametric manova: 1.08850350474611 
#> 
#> Nested anova for Tank nested within Diet 
#> 
#>           Df SumsofSquares      F N.Perm Pr(>F)  
#> Diet       1       0.22595 4.3868    999  0.098 .
#> Tank       4       0.20602 0.9414    999  0.675  
#> Residuals 12       0.65653 0.0547                
#> ---
#> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

^{由reprex package（v0.2.1）于2019-05-16创建}

PRIMER-7的结果： nested PERMANOVA in PRIMER-7

由于随机效应项的变异分量为负，因此将随机效应与主要效应合并起来进行变异划分。以下是带有术语池的嵌套PERMANOVA： nested PERMANOVA with term pooling in PRIMER-7

如何在素食主义者中正确制定嵌套式PERMANOVA？

0 个答案: