Cuffdiff and Cufflink Error with -g option?

[wilson90]

Dear all,

I am using cufflink to produce transcript abundance level estimation, and cuffdiff to do differential analysis.

However, I have come across this contradiction that certain genes, although annotation has been given, is reported as "new" under -g option.

Example,
Entry 1.
XLOC_001015 - - XLOC_001015 Tnnt2 TSS1531,TSS1532,TSS1533,TSS1534 chr1:137732962-137748850 [FPKM values ..]

Entry 2.
XLOC_002303 - - XLOC_002303 - TSS3895 chr1:137732962-137748850 [FPKM values ..]

Quite obvious that these 2 lines in the cuffdiff (genes.fpkm_tracking) file are describing the same gene?

Any plausible explanation?

Thank you.

Best Regards,
Wilson