Illumina sequencing error rates

[sixguns]

Hi, Does anyone know the sequencing error rates in Illumina platform?

[francois.sabot]

The Illumina specificity for the v4 kit (75 PE) is of 1.33 % forward and almost 1.7 reverse, for a Qphred of 30

[sixguns]

The Illumina specificity for the v4 kit (75 PE) is of 1.33 % forward and almost 1.7 reverse, for a Qphred of 30

If my data is 75bp not strand, it means that it close to 1bp error per read?

[francois.sabot]

Yes, theoretically. But it seems to be better to consider it somehow a little higher (around 2% max). Do not forget that you may also have SNP/InDel between your ref and your data, thus explaining some 'errors' you might see

[sixguns]

That's true; Thanks~

[Torst]

If my data is 75bp not strand, it means that it close to 1bp error per read?

The are much more likely to occur at the 3' end of the read, and, a proportion of the reads have much more than one error, so lots of reads will be error free.

HOWEVER I think the Illumina error rate may be misleading us. I believe they only count MAPPABLE reads to their PHIX74 CONTROL. Now, a read which has lots of errors will NOT MAP, hence will not count toward the error rate! Simply by tweaking their mappability thresholds, they can choose any error rate they want to match a reasonable yield...

[francois.sabot]

Yes but how can you calculate the 'true' errors rate, meaning the level of misreading, the level of chimeria, etc... For that you must know perfectly your DNA you are sequencing, and you cannot, due to continual modification of it... However, you can estimate it based on the common sequences between reference and tested sequences, and in calculating the false of non-SNP mismatches /indels.

[cdry7ue]

Francois, is there a paper/document describing in more details the suggestions you just gave above. I have DNA which we prepared by doing PCR on a normal individual,confirmed size by CE and also are planning to do Sanger sequencing.

[francois.sabot]

Francois, is there a paper/document describing in more details the suggestions you just gave above. I have DNA which we prepared by doing PCR on a normal individual,confirmed size by CE and also are planning to do Sanger sequencing.

In fact those specs came from my classical provider (Integragen, France), from their own experience... No paper associated with so. Sorry :s