SEQanswers

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



Similar Threads
Thread Thread Starter Forum Replies Last Post
Power in RNA Seq studies aoifemc RNA Sequencing 0 02-22-2013 08:00 AM
power calculation for exome seq kjaja Bioinformatics 0 12-15-2012 05:30 PM
Power Analysis - Sample Size Calculation jroussarie Bioinformatics 2 11-07-2012 11:15 AM
a question about rpkm calculation camelbbs RNA Sequencing 2 10-16-2012 09:07 AM
Power analysis for RNAseq dglemay RNA Sequencing 0 03-03-2011 08:34 PM

Reply
 
Thread Tools
Old 09-19-2013, 07:48 PM   #1
diviya
Junior Member
 
Location: boston

Join Date: Sep 2013
Posts: 1
Default Power calculation question

I want to compute the power to detect any copy number variant in the genome that is ~100bp, given that I have ~0.5x depth (coverage) of the genome. I was wondering if anyone has suggestions on how to do this power calculation.

Is poisson or beta binomial a better approximation for the coverage of the genome?
diviya is offline   Reply With Quote
Old 09-20-2013, 08:07 AM   #2
thomasblomquist
Member
 
Location: Ohio

Join Date: Jul 2012
Posts: 68
Default

What is the variance in the coverage of your genome? For areas that have at least 12 to 15 reads covering 100 base length area, and if you have a single allelic point mutation, you will have a type I error rate of 0.05 and a type II error rate of 0.20. Approximately. This is based on Poisson sampling laws.

This calculation has a TON of assumptions. For example, it assumes that the CNV (copy number variation) is represented in the entire sample, and not a subpopulation of cells. It assumes that the library preparation effects do not treat the different allelic loci in unique ways. For example, there are different types of CNV, and it may place a copy in a GC rich region which can create ligation chemistry biases that result in abnormal quantification (up to 1000-fold off), which will swamp any signal detection of a 2:1 ratio. This separate region may fragment differently and thus may be detected and ligated differently. If the region fragments are smaller for one of the alleles, it may not cluster efficiently on the Illumina, or may not process properly on the Ion Torrent bead populating step with emPCR.

And these are the biases we are aware of that can effect copy abundance measurement.

A PCR-based prep method that captures a variant region of interest can measure allele ratios quite accurately with vary few reads, but requires a priori knowledge of the region of interest.

Sorry, but this is a very difficult area, and based on my experience a 0.5x coverage is quite low sampling to reasonably detect a significant copy number change to any trust-worthy degree.

Regards, -Tom Blomquist
thomasblomquist is offline   Reply With Quote
Old 09-20-2013, 08:28 AM   #3
thomasblomquist
Member
 
Location: Ohio

Join Date: Jul 2012
Posts: 68
Default

Just to clarify, the above is to detect a 2:1 ratio of alleles.
thomasblomquist is offline   Reply With Quote
Reply

Tags
coverage, depth, power calculation, sequencing

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 07:52 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO