mapping reads to known miRNAs precursors

vebaev · 08-03-2011, 01:15 AM

Dear all,
I'm trying to estimate the all reads count that are mapped to a specific precursor. I have done it my using bowtie with no mistmaches -v 0
But in some papers and tools I see that thay allowed 1 mistmache?! I compared the results in both way. Most of the counts are similar but for some miRNAs thay are doubled like in mir143 bellow:

with 1 mistmach mapping:
miR-143 - 4,697,060
miR-10b - 520,833

with 0 mistmach mapping:
mir-143 - 1,976,574
mir-10b - 501,099

So which is the right way? I thought it should be with 0 mistmaches since non specific tags can align when mistmaches occure but some people do it? And if I want to estimate diff.expression I should know which method is most relevant?

Thanks

DZhang · 08-03-2011, 06:48 AM

vebaev,

a straightforward way to check is to look at the alignments in any genomics viewer. It will tell you exactly which base(s) are not perfectly aligned for these genes. I believe some people allow one mismatch to accommodate sequencing errors or SNPs.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Common elements within two set of miRNAs	Giorgio C	Bioinformatics	1	01-05-2012 10:35 AM
PubMed: Deep sequencing of microRNA precursors reveals extensive 3' end modification.	Newsbot!	Literature Watch	0	08-20-2011 12:02 AM
way to normalize copy number data for small RNAs/miRNAs?	vebaev	Bioinformatics	2	03-28-2011 03:18 AM
Can you really see miRNAs on gels before PCR amplification	fisher	Illumina/Solexa	3	07-27-2010 07:37 PM
PubMed: Prediction of novel miRNAs and associated target genes in Glycine max.	Newsbot!	Literature Watch	0	03-05-2010 03:00 AM

08-03-2011, 01:15 AM	#1
vebaev Senior Res. Location: Plovdiv, Bulgaria Join Date: Oct 2008 Posts: 110	mapping reads to known miRNAs precursors Dear all, I'm trying to estimate the all reads count that are mapped to a specific precursor. I have done it my using bowtie with no mistmaches -v 0 But in some papers and tools I see that thay allowed 1 mistmache?! I compared the results in both way. Most of the counts are similar but for some miRNAs thay are doubled like in mir143 bellow: with 1 mistmach mapping: miR-143 - 4,697,060 miR-10b - 520,833 with 0 mistmach mapping: mir-143 - 1,976,574 mir-10b - 501,099 So which is the right way? I thought it should be with 0 mistmaches since non specific tags can align when mistmaches occure but some people do it? And if I want to estimate diff.expression I should know which method is most relevant? Thanks

08-03-2011, 06:48 AM	#2
DZhang Senior Member Location: East Coast, US Join Date: Jun 2010 Posts: 177	vebaev, a straightforward way to check is to look at the alignments in any genomics viewer. It will tell you exactly which base(s) are not perfectly aligned for these genes. I believe some people allow one mismatch to accommodate sequencing errors or SNPs.