Stringtie take a very long time to compute and never end

[wdesaintjean]

Hello,
I used Stringtie with the basic command on several long reads dataset:

stringtie \\
        -L \\
        -G $gtf \\
        -A ${meta.id}.gene_abund.tab $bam

stringtie \\
        -L \\
        -G $gtf \\
        -o ${meta.id}.stringtie.gtf $bam

But on a long reads dataset of 8 samples, 7 of them compute in 5 ~hours. However, the last sample take most of 60 hours and didn't finish.
All the bam file have approximately the same number of lines (40M ~ 50M) and the problematic sample seem to be perfectly normal: I obtained results with other transcriptome builder like FLAIR.

Did anyone ever had the same problem?

Best regards,
William

[gpertea]

Is this happening with the latest 3.0.0 version? If so, would it be possible to share the $gtf and the BAM file for that sample which gets stringtie stuck ?

[wdesaintjean]

I use the 2.1.4 version. Do you think using the 3.0.0 version can resolve the problem?

To be more precise, stringtie2 create a temporary gtf of 168.393 line from the problematic sample in more than 2 days

wc -l tmp.XXTzG71c/*
168393 tmp.XXTzG71c/WT_P1_R3.stringtie.gtf.tmp

To compare with my 7 other samples, wa have this in less than 5 hours:

   709328 ../../../../results/stringtie2/KO_P1_R1.stringtie.gtf
   817862 ../../../../results/stringtie2/KO_P1_R2.stringtie.gtf
   812990 ../../../../results/stringtie2/KO_P1_R3.stringtie.gtf
   759811 ../../../../results/stringtie2/KO_P1_R4.stringtie.gtf
   816719 ../../../../results/stringtie2/WT_P1_R1.stringtie.gtf
   860879 ../../../../results/stringtie2/WT_P1_R2.stringtie.gtf
   772394 ../../../../results/stringtie2/WT_P1_R4.stringtie.gtf