QuEST found 0 peaks in output

[Yi Li]

Hi,

I did the QuEST analysis but finally found 0 peaks in the output
directories. The command I used was:

generate_QuEST_parameters.pl -QuEST_align_ChIP
foxH1.wt.matched_vs_inputDNA.QuEST -QuEST_align_RX_noIP
foxH1.inputDNA.control.matched.QuEST -gt xenTro72.chrom.sizes -ap
foxH1.wt_vs_inputDNA_QuEST_output

There was an error popped up during running, which said fail to
estimate peak shift, and then asked me choose a peak shift size. I
used the default one, which is 50bp. There are also 3 suggested
parameter sets for calling peaks, which are stringent, suggested, and
relaxed. I've tried the suggested and relaxed version, but both of
them gave me 0 peaks.

There are a lot of information in the final output directory. I am not
sure which may be helpful for you, so I just list some of them.

In the bin_align directory, background, ChIP and pseudo_ChIP
directories are full of .bin files, but the RX_noIP directory is
empty.

###########################################################################################################

In the calls directory, most files are empty files:

-rw-r--r-- 1 yili yili 246 2012-04-23 18:06 peak_caller.ChIP.out
-rw-r--r-- 1 yili yili 0 2012-04-23 18:10
peak_caller.ChIP.out.accepted
-rw-r--r-- 1 yili yili 0 2012-04-23 18:10
peak_caller.ChIP.out.rejected
-rw-r--r-- 1 yili yili 0 2012-04-23 18:10
peak_caller.ChIP.out.with_metrics
-rw-r--r-- 1 yili yili 246 2012-04-23 18:10
peak_caller.pseudo_ChIP.out


###########################################################################################################

In the log directory, the QuEST.log is empty.

###########################################################################################################

In the module_outputs directory.
QuEST.out:
## please cite:
## Valouev A, Johnson DS, Sundquist A, Medina C, Anton E, Batzoglou
S,
## Myers RM, Sidow A
## Genome-wide analysis of transcription factor binding sites based
## on ChIP-Seq data.
## Nat Methods. 2008 Sep; 5:(9):829-35

ChIP peaks: 0
ChIP peaks accepted: 0
ChIP peaks rejected: 0

ChIP regions: 0
ChIP regions accepted: 0
ChIP regions rejected: 0

pseudo_ChIP peaks: 0
pseudo_ChIP regions: 0

ChIP peak FDR estimate: NA
ChIP regions FDR estimate: NA

###########################################################################################################


I am not sure whether this information are useful or not. But if you
have any general suggestions where I did wrong, please tell me.

Thanks very much!

Yi Li

[Yi Li]

BTW, I was running on a xenbase genome, which have thousands of
chromosomes, and most of the chromosomes are pretty short. I'm not
sure whether this might be the reason of 0 peaks found.

[timdeme...@gmail.com]

In line with this topic: using the complete scaffold list I got the same problem, i.e. peak shift estimate = NA. However, after removing the smaller scaffolds (only retaining the 20 or so "genuine" chromosomes), peak shifts could be rather accurately measured.

Best regards,
Tim



Op woensdag 2 mei 2012 06:34:05 UTC+2 schreef Yi Li:

[Federico Gaiti]

Hi all,

I've been trying to run QuEST. I am working on a sponge species I have thousands of scaffolds which a good percentage are relatively short (<10Kb). 

The first trial with QuEST I did not remove the smaller scaffolds and I got peak shift estimate = NA. I then selected the recommended size of 50 and the script went on till the end without errors. 

Though, the QuEST analysis found 0 peaks in the output directories.

This is the command I used:

./QuEST_2.4/generate_QuEST_parameters.pl -sam_align_ChIP /opsin/u/federico/RIP/BAM/Ezh2/Ezh2_sorted_rmdup.sam -sam_align_RX_noIP /opsin/u/federico/RIP/BAM/Input/Input_sorted_rmdup.sam -gt chrom_correct.sizes -ap /opsin/u/federico/RIP/BAM/QuEST/ -ChIP_name Ezh2_RIP

The treatment SAM file looks like this: 

M01177:26:000000000-A8A2V:1:2101:23411:18636 147 Contig9999 2490 50 68M = 2490 -68 CAAACACCACAGTACTCCAATTGCTTGCCGTCCTATACAAGCAATTCAGAGGATAGCAGCTGAAGGAC ?GEFFEE9FGFFGGGGGGFGFF@F:EEGFFGFGGDGE@DADCCF,FGFFF8FE,E,GFEC9FF,CBA- AS:i:-3 XN:i:0 XM:i:1 XO:i:0 XG:i:0 NM:i:1 MD:Z:55T12 YT:Z:UU NH:i:1 XS:A:+

The input SAM file looks like this:

M01177:26:000000000-A8A2V:1:1115:10786:9719 403 Contig9999 3905 0 38M = 3905 -38 CCAATAGTGATTGTACATGTATATCCTTTTGAGTGGGT GGGGGGGFGGGFFEFEGFGGGFGGGFGGGGGGGCCCCC AS:i:0 XN:i:0 XM:i:0 XO:i:0 XG:i:0 NM:i:0 MD:Z:38 YT:Z:UU NH:i:29 XS:A:+ HI:i:28

The chromsome table file looks like this:

Contig13522 1888931
Contig13521 1425252
Contig13520 1092842
Contig13519 1050623
Contig13518 1034607
Contig13517 931383
Contig13516 899346
Contig13515 852899
Contig13514 817803
Contig13513 765180

I was wondering if you guys have any ideas about why I am getting 0 peaks accepted. Will removing the smaller scaffolds fix the issue? 

Thanks and let me know if I should provide additional information for debugging

Federico