I am designing PCR primers to identify human tissues and body fluids via differential mRNA expression. Typically there are multiple known transcripts for each target gene. For example, the blood marker ALAS2 has eight protein-coding transcripts in the Gencode transcriptome and three NCBI RefSeq transcripts. Out of these, NCBI chooses a RefSeq Select transcript (here, NM_000032.5 for ALAS2), which is picked based on multiple criteria such as clinical importance and transcript expression.
Ideally, I would try to design primers targeting all of the known transcripts, but this task becomes increasingly difficult in case of more than two or three transcripts, given that one of the primers should be placed on an exon-exon-junction to avoid amplification of gDNA. However, I believe that not all of the transcripts are equally important. Some transcripts might be generally expressed at low levels. So the question here is, if it's worth it to try to target all of the known transcripts.
Which transcript(s) would you use for primer design?