cole-trapnell-lab / cufflinks

Boost Software License 1.0
310 stars 116 forks source link

max-bundle-frags does NOT work? #97

Closed wangyugui closed 6 years ago

wangyugui commented 6 years ago

Hi.

max-bundle-frags does NOT work?

Command line:
cufflinks -q --no-update-check --num-threads 44 --library-type fr-unstranded --min-isoform-fraction 0.10 --pre-mrna-fraction 0.05 --min-frags-per-transfrag 40 --max-bundle-frags 3813660 --GTF-guide /usr/bio-ref/GRCh38.p11.rnaref/genes.gff --multi-read-correct --output-dir /biowrk/TCM.bio.RNA/gtf.links.GRCh38.p11/pinghe /biowrk/TCM.bio.RNA/gtf.links.GRCh38.p11/pinghe/merge.bam 
[00:57:17] Loading reference annotation.
[00:57:49] Inspecting reads and determining fragment length distribution.
Processed 76347 loci.                       
> Map Properties:
>   Normalized Map Mass: 91893007.95
>   Raw Map Mass: 91893007.95
>   Number of Multi-Reads: 2607507 (with 9616631 total hits)
>   Fragment Length Distribution: Empirical (learned)
>                 Estimated Mean: 247.78
>              Estimated Std Dev: 36.39
[01:19:09] Assembling transcripts and initializing abundances for multi-read correction.

3:47850409-51384198 Warning: Skipping large bundle.

a large bundle reported, but I checked the counts.

#samtools view -c /biowrk/TCM.bio.RNA/gtf.links.GRCh38.p11/pinghe/merge.bam  3:47850409-51384198
1112737

It was small than the param '--max-bundle-frags 3813660'

wangyugui commented 6 years ago

I checked the source, "Skipping large bundle." meas the "--max-bundle-length", not the "--max-bundle-frags".