Min/Max allowable Ns in reads

jfo · 10-16-2016, 07:21 AM

Dear all,

How much Ns in a read should be allowed before discarding it? I have read a few studies removing reads with >1 N; but most studies on RNA-seq do not mention about this parameter. Is this too stringent for de novo assembly and differential expression studies using Illumina Hiseq 100 PE? All suggestions will be very much appreciated. Thank you!

GenoMax · 10-16-2016, 02:40 PM

N's appears in reads due to the inability of the basecaller to assign a specific base. In generally they should be rare/absent. You may see them in libraries with low nucleotide diversity. You obviously don't want them in data that would be de novo assembled. If they are present at the beginning/end of the reads and can be easily trimmed then you should remove them.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Using bowtie2 with min and max fragment fragment	millrick10	Bioinformatics	12	05-27-2015 07:03 PM
min length and max length with fastx tookit	Sergio.pv	Bioinformatics	6	11-19-2013 01:53 PM
CLC paired read orientation- max. and min. read distance	gevielr	Bioinformatics	1	11-10-2013 03:56 PM
calcultate Max, min of contigs length	Votinhkiem90	Bioinformatics	6	09-30-2013 04:54 PM
min-coverage-normal & min-coverage-tumor in VarScan somatic	Jane M	Bioinformatics	2	11-20-2012 09:46 AM

10-16-2016, 07:21 AM	#1
jfo Junior Member Location: SE Asia Join Date: Oct 2016 Posts: 5	Min/Max allowable Ns in reads Dear all, How much Ns in a read should be allowed before discarding it? I have read a few studies removing reads with >1 N; but most studies on RNA-seq do not mention about this parameter. Is this too stringent for de novo assembly and differential expression studies using Illumina Hiseq 100 PE? All suggestions will be very much appreciated. Thank you!

10-16-2016, 02:40 PM	#2
GenoMax Senior Member Location: East Coast USA Join Date: Feb 2008 Posts: 7,087	N's appears in reads due to the inability of the basecaller to assign a specific base. In generally they should be rare/absent. You may see them in libraries with low nucleotide diversity. You obviously don't want them in data that would be de novo assembled. If they are present at the beginning/end of the reads and can be easily trimmed then you should remove them. Last edited by GenoMax; 10-16-2016 at 02:43 PM.