paired-end vs. mate-pair

[lre1234]

Hi all,
we are planning on doing some whole genome sequencing and I have a question. I have been under the impression that paired-end sequencing is the best method to use, but now some of the people here are telling me that we should use mate-pair.

While I know the differences between the two methods, I am not sure which is the best one to use. Does anyone have any thoughts on which one is better, or know of a pub that has compare the two methods?

Thanks for your help and thoughts.

[HESmith]

Mate-pair is a specific type of library; paired-end is a type of sequencing. They are not two different methods; in fact, mate-pair libraries require paired-end sequencing.

The decision to use mate-pair vs. standard libraries depends upon your application.

[swbarnes2]

Mate pair allows you to have your pairs be much farther apart, which can be more informative than the standard paired-end protocol.

The downside is that it's a more complicated protocol, and it can be contaminted with ordinary paired-end reads, which I image can make analysis more difficult.

[krobison]

The other class of challenge from mate pair data are false mates -- cases in which the ligation step brings two unrelated molecules together. The panda genome paper is an example where this was a very serious issue.

[cw11]

This site describes some of the potential restrictions of the mate-pair library. I also found in this thread:

I think paired end refers to sequencing from the ENDS of a DNA fragment. If you're doing 75-bp paired end read using an insert size of 300bp, then the machine will sequence 1-75 and 300-225 (for simplicity, omitting the adapters)

Mate pair requires a completely different protocol and is typically over longer distances such as 2-5kb. If you want to sequence a 5kb mate pair library, then 5kb fragments of DNA are isolated on the gel, the ends are biotinylated, the fragment is circularized and sheared. So now when you select using streptavidin, you'll get the fragment that has the ENDS of the original 5kb fragment. This fragment is then sequenced.

Mate pair is more relevant in genome assembly, especially for covering repetitive sequences. Paired end can be used for anything - RNA, DNA.

You can get more information on the Illumina site:

http://www.illumina.com/technology/paired_end_sequencing_assay.ilmn

http://www.illumina.com/technology/mate_pair_sequencing_assay.ilmn

I'm not sure how 454 or other methods define these terms

[meganathan.pr]

Hi all,
I am trying to prepare mate pair library. I dont wanna use Illumina kit as it is costlier. Do any one have the protocol using other chemical supplies? can I get it?

Thanks

[masker]

I remember that because of the single read length, the 454 defined "paired-end" as "mate-pair" from Illumina.

[SEQond]

I find Illumina's explanation for paired-end quite lacking. Wikipedia has a better explanation

[chevrm]

Mate-pairs are wonderful if you're dealing with particularly tricky assemblies, CNVs, or looking for other structural variations (translocs, etc) but may not be worth it if you don't have any of these criteria.