P value or P adj value

Question

[Claire]

Hello, p value or p adj value is better ? If i have one replicate the result of p adj value will change ?

Thank you

Answer 1

[Kevin Blighe]

Hi Claire, if you have derived a p-value from a comparison of 1 sample versus 1 other sample, then this p-value is virtually meaningless. The same is true for the adjusted p-value.

For further information, please elaborate on your experimental design and how you have processed your data, and also [elaborate] on which statistical test(s) you have employed.

Kevin

[Claire]

Thank you kevin for your kind reply. I did the mean of each sample so now i have 18 samples instead of 54. i am trying to plot heat map correlation between species and chemical data.( i have 2 matrices, Species data and chemical data for each sample) What i want to do it is to show with a star the sig difference between correlation. The problem that padj value is 1 in almot all the correlations. Package ( microbiome)

  y <- log10(abund_table) 
    x <- as.matrix(meta_table)
    correlations <- associate(x, y, method = "pearson", mode = "matrix", p.adj.threshold = 1, n.signif = 1)
    p <- heat(correlation.table, "X1", "X2", fill = "Correlation", star = "p.adj", p.adj.threshold = 1) 
    print(p)