I have a bunch of gene names of Apis mellifera (specifically 194). I used these gene names as an input on STRING database to create a network for Drosophila melanogaster. 54 of those genes were also present in D. melanogaster and I received a second network. Is this a credible method of creating homologous gene network. Or should I use the sequences of the 194 genes and BLAST it to the geneset of the D. melanogaster and do it that way?
Gene names are not consistent across species. Most of the time genes are automatically annotated and the algorithms try to give names that are similar to their matches in other organisms, but not always. It's also more likely that genes will be automatically named/annotated based on their sequence, and manually named/annotated based on their function.
The further away you get from Homo sapiens & Mus musculus (and maybe Caenorhabditis elegans, Arabidopsis thaliana & Danio rerio), the less likely it is to find consistently named genes.