cufflinks -skipped fragment

[capricy]

I align RNA-seq reads to genome and then used cufflinks to do the assembly. But the output always skipped some regions for the gene call even though by IGV I could tell good alignments are existing there.


Here are the result lines from: genes.fpkm_trackingCUFF.1 - -
---------------------------
...

CUFF.10 - - CUFF.10 - - Scaffold1:111637-112521 - - 17.8444 13.9032 21.7856 OK
CUFF.11 - - CUFF.11 - - Scaffold1:114932-115447 - - 5.29068 2.10028 8.48108 OK
CUFF.12 - - CUFF.12 - - Scaffold1:117626-118329 - - 10.7109 7.14059 14.2812 OK
CUFF.13 - - CUFF.13 - - Scaffold1:116233-117155 - - 14.4199 10.9729 17.867 OK
CUFF.14 - - CUFF.14 - - Scaffold1:105973-107198 - - 76.2951 69.6545 82.9358 OK
CUFF.15 - - CUFF.15 - - Scaffold1:112755-114371 - - 93.4633 87.2185 99.708 OK
CUFF.16 - - CUFF.16 - - Scaffold1:114474-114851 - - 108.774 89.2379 128.311 OK
CUFF.17 - - CUFF.17 - - Scaffold1:27634-102622 - - 5567.72 5540.51 5594.93 OK
CUFF.18 - - CUFF.18 - - Scaffold1:618359-618942 - - 3.93266 1.44543 6.41989 OK
...
---------------------------
The genes I am interested in are located at around 300k, but here there are only genes before 117k and after 618k

What should I do? Thanks

[capricy]

by the way the gene alignment was done by tophat

[capricy]

I feel it might be related to the following parameters. Anyone can give a hint about how to set it? Thanks.


--max-bundle-frags

[severin]

# of reads

You have too many reads on that fragment so it skips it.

You can filter your alignment file so that there is no more than say 20 reads or 100 reads at any given base and that should fix your issue. I have not found a way to set the parameters so that it will work.

[capricy]

Here is my warning message. How should I set this?

------------------
Warning: Using default Gaussian distribution due to insufficient paired-end reads in open ranges. It is recommended that correct parameters (--frag-len-m
ean and --frag-len-std-dev) be provided.