I would like to compare homologus genes in different species and I do not want to rely on experimental data for gene expression. I was searching the literature and I found thet computing gCAI (which derives a set S from genome) and the "effective usage of codons - Nc" which does not need any reference set to calculate it. I am wondering if anyone have experiences about which one is better to use if I would like to compare a codon usage between two different species that have different codon usage bias.

Thank you for your answers! Martin