Understanding the gffcompare output for my dataset

Question

[amy16]

Hi all, Below is the output from gffcompare of assembled transcripts from my RNA-seq experiment with the reference genome annotation. I am concerned about the low precision values. Do I have to rerun my assembly?

Query mRNAs : 74136 in 32930 loci (63219 multi-exon transcripts)

(13176 multi-transcript loci, ~2.3 transcripts per locus)

Reference mRNAs : 28269 in 28269 loci (20429 multi-exon)

Super-loci w/ reference transcripts: 27162

-----------------| Sensitivity | Precision |

    Base level:   100.0     |    65.6    |
    Exon level:   100.0     |    58.8    |
  Intron level:   100.0     |    72.2    |

Intron chain level: 100.0 | 32.3 | Transcript level: 100.0 | 38.1 | Locus level: 100.0 | 83.2 |

 Matching intron chains:   20429
   Matching transcripts:   28269
          Matching loci:   28269

      Missed exons:       0/139417  (  0.0%)
       Novel exons:   44143/240416  ( 18.4%)
    Missed introns:       0/111148  (  0.0%)
     Novel introns:   22898/153871  ( 14.9%)
       Missed loci:       0/28269   (  0.0%)
        Novel loci:    5421/32930   ( 16.5%)

Total union super-loci across all input datasets: 32930 74136 out of 74136 consensus transcripts written in prrmerg07.annotated.gtf (0 discarded as redundant)

Thanks

[Morgan S.]

Sorry, I know it has been a long time since this was posted, but I also have low precision values and I am wondering if you found out why that is?