General mapping rate of human resequencing data against reference in GAiix/Hiseq

cybog337 · 01-11-2011, 11:23 PM

Hi GA/Hiseq users!

I would like to know general mapping rate of human reseq data produced by Hiseq/GA against human reference.
Some site says 95% mapping of >q30, however another site says under 80% or under 75%. I guess what mapping rate is usually acceptable.

Thankm you!!

simonandrews · 01-12-2011, 07:54 AM

This will depend on the quality of your data and the length of your reads. We see somewhere over 70% mapping from a good 40bp run, but some of the unmapped stuff will be parts which aren't in the assembly. 95% mapping sounds very optimistic even for very long reads with good quality.

For comparison PhiX maps at around 98.5% which suggests that the reduction in human is due to genome complexity and not sequence quality.

cybog337 · 01-12-2011, 09:43 AM

Dear Simon Andrews,
Thank you very much for the reply.
We usually read 100PE using GAIIX or HIseq2000 and mapping rate is 70~80% againt human. Depending your advice, this mapping rate is likely to be normal.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Quality control of genomic resequencing data from a HiSeq	gavin.oliver	Genomic Resequencing	2	06-30-2013 02:48 AM
Help to analyze Illumina HiSeq 2000 Human data	kiradi	Bioinformatics	4	12-09-2011 05:30 AM
High phasing on GAIIx and HiSeq	james hadfield	Illumina/Solexa	3	05-27-2011 07:04 AM
Minimal fragment size for GAIIx or HiSeq??	Unununium	Illumina/Solexa	3	05-02-2011 04:45 AM
GAIIx versus HiSeq Libraries	NextGenSeq	Sample Prep / Library Generation	2	07-22-2010 07:39 AM

01-11-2011, 11:23 PM	#1
cybog337 Junior Member Location: RTP Join Date: Jul 2010 Posts: 8	General mapping rate of human resequencing data against reference in GAiix/Hiseq Hi GA/Hiseq users! I would like to know general mapping rate of human reseq data produced by Hiseq/GA against human reference. Some site says 95% mapping of >q30, however another site says under 80% or under 75%. I guess what mapping rate is usually acceptable. Thankm you!!

01-12-2011, 07:54 AM	#2
simonandrews Simon Andrews Location: Babraham Inst, Cambridge, UK Join Date: May 2009 Posts: 871	This will depend on the quality of your data and the length of your reads. We see somewhere over 70% mapping from a good 40bp run, but some of the unmapped stuff will be parts which aren't in the assembly. 95% mapping sounds very optimistic even for very long reads with good quality. For comparison PhiX maps at around 98.5% which suggests that the reduction in human is due to genome complexity and not sequence quality.

01-12-2011, 09:43 AM	#3
cybog337 Junior Member Location: RTP Join Date: Jul 2010 Posts: 8	Dear Simon Andrews, Thank you very much for the reply. We usually read 100PE using GAIIX or HIseq2000 and mapping rate is 70~80% againt human. Depending your advice, this mapping rate is likely to be normal.