Hi everyone,

I am new to R and have been tasked with analyzing a dataset that only provides FPKM (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE223176). I know that FPKM is not meaningful for sample-wise comparison but it seems they only give the FPKM data.

I have been reading threads regarding this topic but could not figure out a detailed pipeline and input code that I need to use to solve the task. All that I know is it is better to transform the data to log2(FPKM+0.1) and apply limma-trend. However, when I try to follow the user guide of Limma, it seems they only use the integral-number-count-table.

Could someone help me by providing a source that I can read or some example code? I truly appreciate any help since I have no clue what should I do now...

limma-trend can accept FPKM on log scale but it's not optimal, see Differential expression of FPKM from RNA-seq data using limma and voom()

Since this is only eight samples you could use the sra-explorer.info website to get download links for the fastq files and then do a quantification with salmon as described here: https://master.bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html Basically you could follow the linked workflow, and then decide whether you want to use edgeR, limma, DESeq2, they all are established and do well.