[Useq-users] 0 read peaks with ScanSeqs

[Noboru Jo S. <ns...@uc...>]

     Hi David, I'm using ScanSeqs to find peaks in RNA-seq data using 
peaks shift = 100 and window size = 50 and I noticed that a considerable 
number of the peaks found are very short (<20 bp) and have 0 read. 
ScanSeqs reports BW_Sum  BW_Sum+ BW_SumT- such as 2.0 0 2.0, but when I 
overlap the peak with my reads, I find none.

     Do you know what I could be doing wrong?


java -Xmx7500M -jar mypath/USeq_7.6.9/Apps/ScanSeqs -t all_Point/ -s 
results -p 100 -w 50

java -Xmx6000M -jar 
~/shared/programs/chip/USeq_7.6.9/Apps/EnrichedRegionMaker -f 
results/windowData50bp.swi -i 0 -s 0

     I used -i 0 and -s 0 because I couldn't find information about the 
index number for BW_Sum and I actually wanted all the peaks so that I 
could filter myself later.

     If you want to have a look at my .bed and .xls files, I put them at
http://mnlab.uchicago.edu/download/


     The last peak in the .xls file is only 19bp long and doesn't 
overlap any read.

     I can postprocess the peaks, and filter by minimum number of reads, 
but I'm curious whether I'm doing something wrong.

     Thanks!