Findpeaks Fp4 Command For Sam Files

Question

[suman]

Hi, I have recently downloaded Findpeaks4.0 for peak analysis from bowtie outputs e.g. sam files. I need an example command if somebody could kindly provide it.

the test cmd is not woking:

java -jar FindPeaks.jar -input -aligner 22.test.eland -output  C:/ -eff_frac  0.7 -dist_type 1 200  -hist_size 200

the error msg :aligner format 22.test.eland is unknown how to change the cmd?

E.G. I found a chipseq cmd in the forum:

java -jar FindPeaks.jar -name xyz -dist_type 1 200 -minimum 8 -hist_size 100 -trim .2 -subpeaks .5 -iterations 1 -eff_frac 0.7 -output /test -input sample.sam -compare control.sam -aligner sam

will the above cmd be suitable for mouse mm9 aligned reads? adding java -Xmx3G since I have 4GB ram only will help?

sumon

Answer 1

[Sukhi Singh]

The input is missing. I just checked the test command, its

java -jar FindPeaks.jar -input 22.test.eland -output /output/directory/ -eff_frac 1.0 -dist_type 1 200 -directional -filter -hist_size 200

It will be fine for mm9, just change the format for file that you are inputting in the aligner parameter. Read Wiki for more examples.

Answer 2

[suman]

the above reply from sukhdeep gave me the error :aligner type must be provided with the -aligner flag also directional and filter flags could not be processed.; since it is not taking eland as default.So,i had to add also -aligner eland and remove directional filter from the cmd Maybe earlier versions reqd these flags.

changing these gave two regions as peaks and a wig.gz file to my output directory.

The peaks file have the format : idchromstartendmax_coord;

so i guess now its ok. will try the sam files to check it out and post any problems which arise.

thanks for the info. sumon.

Answer 3

[suman]

Hi,

I had two bam files oct4 and gfp (mapping to only chr1 of mm9 genome) which I wanted to analyze with FindPeaks 4.0.Setting -Xmx2G always java VM failed but i have 4gb ram windowsxp(32bit), maybe it cant allocate even half. so I used the cmd:

java-Xmx 1500m -jar FindPeaks.jar -name xyz -disttype 1 200 -subpeaks 0.2 -bedgraph -efffrac 0.7 -output C:/ -input chip.bam -compare control.bam -aligner sam

I got some png files regions graph as output viz: pregraph,first graph,second graph with lines drawn are they regression lines with equations how do interpret them? plus wig files (wig files one for ctrl and one for sample) , but only one bedgraph file?

Which file in the following should I use as the peak file among the following:

a) xyz triangle subpeaks with id,chrom,start,end and max coord or b) xyz traingle subpeaks filtered sample peaks file with id,chrom,start,end,max cord, height,prob sequence

Could anybody elucidate slightly or suggest further tuning parameters to improve the output

sumon

[Ram]

Hi, As I am also using Findpeaks 4.0 with sam files, but it is executing forever. So is it possible to know is there any error or why its not finishing?? It would be quite helpful if you throw some light on it.

Thanks a lot.

Answer 4

[suman]

Hi, I have a question regarding the -iterations and -landerwaterman options. I read that the landerwaterman options best with disttype 0 as it performs poorly with weighted (-disttype 1) distributions and the native/PET (-dist_type 3) distributions.

a)Is the value of integer following dist_type 0 the tag length like 26 reads or 36 reads that we provide in MACS ? b)Can i use both options together in a command ?

suman