Pathway analysis is generally performed using the genes (transcripts) themselves. It is good to keep in mind that the perturbation of gene expression does not necessarily result in changes in the amount of protein. For interpretation purposes, this aspect is usually simplified and the genes are treated as if they were corresponding to the proteins.

There are multiple tools available for pathway analysis and many of them take multiple annotation types as input. The tool then does the conversion for you if needed. However, you may need to convert the transcripts to genes. You can do this with BioMart (https://www.ensembl.org/info/data/biomart/index.html, or BioMart package on R), for example.

Notice that if you are dealing with transcriptomics data and you have fold changes or other values for each transcript, these need to be either collapsed for the transcripts mapping to the same gene entity or otherwise dealt with. Doing this with fold changes and p-values is not the best choice, but instead I would advice to do this prior to differential expression analysis.

I hope this helps.

Use https://anaconda.org/bioconda/transdecoder to convert your transcriptome sequence or find the potential coding areas in your transcript sequences.

Hi Mousumi Mary

as Laura said there are many ways for analyzing your data without transforming transcriptomics into proteomics, you can use IPA (with conservation of up and down regulation) but it needs a licence, or other free tools as DAVID, GSEA, StringDB...

all the best

fred

Thanks, everyone.