IsoQuant extended annotation missing transcripts

[jcmtnez]

Hello,

Thank you for making such a useful tool! When running IsoQuant with a reference GTF file, the final extended GTF contains less transcripts than the original reference despite discovering novel transcripts. I just wanted to make sure this is not related to the parameters I use run the program.

My code:
isoquant.py --reference /work/users/j/c/jcmtnez/index/GRCh38_gencode_v45/GRCh38.primary_assembly.genome.fa
--genedb /work/users/j/c/jcmtnez/index/GRCh38_gencode_v45/transcripts.gtf --complete_genedb
--data_type pacbio_ccs
--fastq /work/users/j/c/jcmtnez/RNAseq/SF_long/fastq/WT.fastq
/work/users/j/c/jcmtnez/RNAseq/SF_long/fastq/SF3B1.fastq
/work/users/j/c/jcmtnez/RNAseq/SF_long/fastq/SRSF2.fastq
/work/users/j/c/jcmtnez/RNAseq/SF_long/fastq/U2AF1.fastq
--sqanti_output
-o /work/users/j/c/jcmtnez/RNAseq/SF_long/isoquant/

The result from GFF compare:

gffcompare v0.10.4 | Command line was:

#gffcompare -r transcripts.gtf OUT.extended_annotation.gtf

#= Summary for dataset: OUT.extended_annotation.gtf

Query mRNAs : 206377 in 31488 loci (191408 multi-exon transcripts)

(16483 multi-transcript loci, ~6.6 transcripts per locus)

Reference mRNAs : 251242 in 57581 loci (225209 multi-exon)

Super-loci w/ reference transcripts: 25892

#-----------------| Sensitivity | Precision |
Base level: 72.0 | 94.8 |
Exon level: 77.1 | 97.3 |
Intron level: 76.9 | 97.3 |
Intron chain level: 78.5 | 92.4 |
Transcript level: 75.1 | 91.5 |
Locus level: 52.9 | 96.0 |

 Matching intron chains:  176850
   Matching transcripts:  188791
          Matching loci:   30477

      Missed exons:  151359/666158	( 22.7%)
       Novel exons:    4550/535946	(  0.8%)
    Missed introns:   91276/402509	( 22.7%)
     Novel introns:    1336/318070	(  0.4%)
       Missed loci:   25270/57581	( 43.9%)
        Novel loci:     847/31488	(  2.7%)

Total union super-loci across all input datasets: 31483
206377 out of 206377 consensus transcripts written in gffcmp.annotated.gtf (0 discarded as redundant)

Thank you for your time!
Jose

[andrewprzh]

Dear @jcmtnez

Thank you for the positive feedback!

Thanks for the report. Yes, this is a known bug related to extended annotation only. It is now fixed and the bug-fix release will be out soon.

Best
Andrey

[andrewprzh]

Finally released new version 3.4, which fixes this issue.