I have 18 samples. These samples have 2 different factors with one factor(Infestation) having 2 levels (as control vs affected) & another factor(timepoint) having 3 levels(as 24h, 48h,96h) . Also the last factor (timepoint) have 3 biological replicates.
counts <- read.delim("counts.csv", header = TRUE, row.names =1, sep = ",") dim(counts)  27868 18 colData <-read.delim("colData.csv", header =TRUE, sep = ",", row.names =1) dds <- DESeqDataSetFromMatrix(countData = counts, colData = colData, design = ~ Infestation+timepoint) dds <- DESeq(dds) res <- results(dds) de_genes <- rownames(res)[which(res$padj < 0.5 & abs(res$log2FoldChange) > 1)][1:50]
This piece of code is getting very less genes( less than 10) using padj & log2fold values above mentioned.
BUT, modifying my design formula to include only one factor(i.e., Infestation) while disregarding timepoint factors gives me my specified number of 50 genes.