Cufflinks returning elevated FPKM values for 'new' transcripts

[gwilymh]

I am running RNA-Seq in a Tuxedo pipeline. In the final Cufflinks output, I am getting massively elevated FPKM values for some transcripts. These inflated values are only returned for the transcripts newly discovered by Cufflinks/TopHat (i.e. transcripts that were not previously annotated). The transcripts from annotated genes seemed normal.

In one analysis, for example, the known genes had an average FPKM value of 214.5, and a maximum FPKM value of 147,473. The newly discovered transcripts, however, returned a mean FPKM of 139,234.6 and a maximum of 74,769,200; 2.5% of the new transcripts had vales greater than the maximum FPKM detected for the annotated genes. The results for the new transcripts clearly contain artifacts.

My pipeline is below
tophat \
--min-anchor-length 10 \
--splice-mismatches 1 \
--min-intron-length 5 \
--microexon-search \
--fusion-search \
--transcriptome-index=/gpfs/group1/f/flyinv/working_index/Dpse3_0_1_exons_ \
-o "/gpfs/group1/f/flyinv/Outputs_TopHat/transcriptomeSequence_exons/AR_MSH126_Male" \
/gpfs/group1/f/flyinv/working_index/Dpse3_0_1 \
/gpfs/group1/f/flyinv/RNASeq/AR_MSH126_Male_1_sequence.txt \
/gpfs/group1/f/flyinv/RNASeq/AR_MSH126_Male_2_sequence.txt \

cufflinks \
--output-dir "/gpfs/group1/f/flyinv/Outputs_CuffLinks/transcriptomeSequence_exons/AR_MSH126_Male" \
--GTF-guide /gpfs/group1/f/flyinv/working_index/Dpse3_0_Exons.gff3 \
--upper-quartile-norm \
--min-intron-length 5 \
--quiet \
/gpfs/group1/f/flyinv/Outputs_TopHat/transcriptomeSequence_exons/AR_MSH126_Male/accepted_hits.bam \


Does anyone know how or why this occurs? How can it be prevented? How can such artifacts be screened out of downstream analyses?