Discrepancy in Quantification Results with and without -e Option

[HadoBoirudoWandarando]

Hello,

I have encountered an issue with StringTie's quantification results that vary based on the use of the -e option. Specifically, I have observed a difference in the abundance file outputs for a gene when running StringTie with and without the -e option.

When I run StringTie with the -e option, the output in the abundance file shows zero coverage, FPKM, and TPM values for a gene (thrL) that has existing annotations:

Gene ID    Gene Name    Reference    Strand    Start    End    Coverage    FPKM    TPM
gene-b0001    thrL    NC_000913.3    +    190    255    0.000000    0.000000    0.000000

However, when I run the same command without the -e option, the output shows non-zero values for the same gene:

Gene ID    Gene Name    Reference    Strand    Start    End    Coverage    FPKM    TPM
gene-b0001    thrL    NC_000913.3    +    190    255    452.000000    158.843719    97.874275

Despite this, the corresponding GTF file still shows zero values:

NC_000913.3    StringTie    transcript    190    255    1000    +    .    gene_id "sample1.1"; transcript_id "sample1.1.1"; reference_id "gene-b0001"; ref_gene_id "gene-b0001"; ref_gene_name "thrL"; cov "0.000000"; FPKM "0.000000"; TPM "0.000000";
NC_000913.3    StringTie    exon    190    255    1000    +    .    gene_id "sample1.1"; transcript_id "sample1.1.1"; exon_number "1"; reference_id "gene-b0001"; ref_gene_id "gene-b0001"; ref_gene_name "thrL"; cov "0.000000";

Could you please help me understand why there is a variation in the quantification results for a gene with existing annotations, depending on whether the -e option is used or not? Additionally, why does the GTF file still show zero values despite the non-zero values in the abundance file when -e is not used?

Thank you for your assistance.