cufflinks confusing

[MeixiaZhao]

I have some confusings about cufflinks output file:
1. Why the cufflinks output can match the isoforms but can not match the gene, like the first one:

Gm01 Cufflinks transcript 51481 61502 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; FPKM "0.4311150338"; frac "1.000000"; conf_lo "0.045514"; conf_hi "0.816716"; cov "0.409373"; full_read_support "no";
Gm01 Cufflinks exon 51481 51625 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; exon_number "1"; FPKM "0.4311150338"; frac "1.000000"; conf_lo "0.045514"; conf_hi "0.816716"; cov "0.409373";
Gm01 Cufflinks exon 51828 51925 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; exon_number "2"; FPKM "0.4311150338"; frac "1.000000"; conf_lo "0.045514"; conf_hi "0.816716"; cov "0.409373";
Gm01 Cufflinks exon 52362 52465 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; exon_number "3"; FPKM "0.4311150338"; frac "1.000000"; conf_lo "0.045514"; conf_hi "0.816716"; cov "0.409373";
Gm01 Cufflinks exon 52517 52571 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; exon_number "4"; FPKM "0.4311150338"; frac "1.000000"; conf_lo "0.045514"; conf_hi "0.816716"; cov "0.409373";
Gm01 Cufflinks exon 52608 52639 1000 - . gene_id "CUFF.1"; transcript_id "PAC:16242891"; exon_number "

gene_id "CUFF.1"; transcript_id "PAC:16242891"; PAC:16242891 in the reference annotated file is the transform of gene Glyma01g00270.1? Can I recognise this gene CUFF.1 as Glyma01g00270, and those two transcripts belong to this gene?

2. I have noticed that many transcripts' FPKM value is 0.0000000000, should I exclude those transcripts? I suppose not. I just wondering wheather I need to set a FPKM cutoff after the output of cufflinks. Actually, even some transcript FPKM value is 0, it also can be significant expressed when using cuffdiff.