SEQanswers

Go Back   SEQanswers > Sequencing Technologies/Companies > Illumina/Solexa



Similar Threads
Thread Thread Starter Forum Replies Last Post
HiSeq mystery: bad quality scores affecting mainly T and G in first four cycles sfbiologist Illumina/Solexa 6 09-10-2012 10:39 AM
Has any one here used 200 cycles SBS kit to perform reads shorter than 50 cycles? Lanlan Illumina/Solexa 2 02-27-2012 08:54 PM
How many cycles do you *really* get? jlove Illumina/Solexa 3 06-24-2011 04:32 AM
Drop in scores after 41 cycles mvignali Illumina/Solexa 1 07-30-2010 12:08 PM
How many cycles you can get out of the 36 cycle kit version 2? schönblick Illumina/Solexa 12 07-22-2009 05:34 AM

Reply
 
Thread Tools
Old 08-28-2013, 03:11 AM   #1
seqask
Junior Member
 
Location: Barcelona

Join Date: Aug 2013
Posts: 2
Question 2x150 cycles vs 2x250 cycles: coverage, quality, aligning and cost issues

Hello,
I am planning a targeted resequencing experiment using the Illumina TruSight Cancer panel, which claims an average insert size of 300bp. Illumina people recommend MiSeq 2x150 runs but, as in my sequencing service MiSeq 2x250 runs are just 8% more expensive than MiSeq 2x150 runs, I wonder if it wouldn't be worth to use MiSeq 2x250 runs. I understand that many pair reads will overlap, but I still think I will gain a considerable amount of coverage. I'd like to find medium (1-50bp) insertions and deletions and hope that longer reads will help me to align them properly. Is this correct? Are any quality differences between those chemistries that I should take into acount? Any appreciation is welcome!
seqask is offline   Reply With Quote
Old 08-28-2013, 10:05 PM   #2
luc
Senior Member
 
Location: US

Join Date: Dec 2010
Posts: 343
Default

I have no experience with trueSight protocols, but it seems you might get higher quality data due to the huge overlaps when following the standard procedure and using a read merger like Flash afterwards. If you modify the tagmentation reaction (more input DNA?) or do a size selection before the PCRs you should be able to actually sequence longer fragments.
luc is offline   Reply With Quote
Old 08-29-2013, 11:40 PM   #3
seqask
Junior Member
 
Location: Barcelona

Join Date: Aug 2013
Posts: 2
Default

Thank you Luc! If I understand correctly, you mean 2x250 cycle runs can be worth. After the pair read merging with Flash, will I be able to use merged and unmerged reads together in an aligner like BWA-MEM or so?
Regarding to the insert sizes, I think that the TruSight oligo and tagmentation kits are matched to provide a probe distribution adapted to the fragment lenghts, so I would like to use the system "as is" first, and just trying to get the maximum coverage and the longer fragments for accurate aligning and variant calling.
seqask is offline   Reply With Quote
Old 08-31-2013, 07:27 AM   #4
mcnelson.phd
Senior Member
 
Location: Connecticut

Join Date: Jul 2011
Posts: 162
Default

The data quality for a 2x250 run will be lower overall than for a 2x150 run, that much is a given. This will mostly be the result of lower q-scores at the 3' end of the reads, and for some 2x250 runs the drop-off in quality after ~150 bp can be quite dramatic.

Now, as luc said, with longer reads you could probably overlap the ends of the read pairs to form single reads with a higher average quality. Depending on what programs you'll be using for variant calling or the like, that can be a good or a bad thing. The reason is that while your reads will be longer and of potentially higher quality, which can help you align them to your reference more accurately, you'll have lost read depth. When a read pair overlaps, a number of variant callers treat that as only a single count versus two because both reads originated from the same library fragment and thus a change could be the result of a PCR artifact or similar. Thus, you would need more reads to get the same level of sensitivity as a library where the reads don't overlap.

Having said all that, if the extra cost of a 2x250 run isn't that big of a deal for you, then go ahead and do it. That way you get the longer reads which may prove more useful, and the worst case scenario is that you end up trimming the data to 150bp because the longer reads lengths cause too many issues.
mcnelson.phd is offline   Reply With Quote
Reply

Tags
cost, efficiency, indel aligning, read lenght

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 01:42 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO