Cufflinks shows novel and abundant transcript

Question

[nand]

I ran cufflink with -g option using gencode annotion file gencode.vM15.chr_patch_hapl_scaff.annotation.gtf. I sorted *the isoforms.fpkm_tracking based on fpkm and found that there many CUFF.XXX transcripts that have very high FPKM. The deep sequencing data corresponds to a paper published in 2014 September. So, if the CUFF.XXX transcripts are very abundant in a 2014's experiments, shouldn't those transcript be already there in the gencode annotation file and have a transcript id like ENSMUST00000XXXXXX?

These are the top transcripts:

[GenoMax]

shouldn't those transcript be already there in the gencode annotation file

Not necessarily. Official recognition of a transcript requires experimental evidence. Here is the model that Ensembl uses.