Question

Best way to do quality control and analysis on Exon arrays

1

Entering edit mode

Daniel Brewer ★ 1.9k

@daniel-brewer-1791

Last seen 9.6 years ago

Hi, I have played around with small Exon 1.0 ST datasets before and have been generally been using affymetrix power tools (APT) to do the summarisation and then the Affymetrix annotation using the "core" set. I am about to start on a larger set of analysis and so I am revisiting what are the best tools to use. There now seems to be three main alternatives with R to do the summarisation and one outside R. A couple of questions: 1) What is the best way to do quality control on these arrays? I was thinking about looking at the spread of DABG p-values but for exonmap at least it doesn't seem like you can calculate these values. Can you use AffyQCreport? 2) If I decided to use APT for the summarisation are there annotation packages available? exonmap seems to be have an annotation but this seems to be redone, is there any based on the affy supplied annotations and the "core" and "extended" groups? Thanks Dan -- ************************************************************** Daniel Brewer, Ph.D. Institute of Cancer Research Molecular Carcinogenesis Email: daniel.brewer at icr.ac.uk ************************************************************** The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the a...{{dropped:2}}

Annotation Cancer affy exonmap Annotation Cancer affy exonmap • 1.3k views

ADD COMMENT • link updated 8.7 years ago by nicholaserho ▴ 10 • written 14.8 years ago by Daniel Brewer ★ 1.9k

score 1 · Answer 1 · 2015-08-05

There are two quality control metrics that can be calculated fairly easily using APT. The first you already mentioned is using the detection above background algorithm (DABG). While it is typically not a good idea to apply this to gene level (i.e. core, extended, etc.) you can calculate it for every probeset using APT. Once these detection p-values are calculated you can compute the percentage of the chip detected by applying a p-value threshold. In my experience you want to have around a 30% detected rate or above. The quality of the chip will translate to gene level.

Positive vs negative area under the curve (AUC) is the second basic quality control metric. APT will also calculate this for each chip in this calculation APT is assessing the discrimination of probes which should be "on" (i.e. house keeping genes, etc) compared to probes which should have no hybridization (i.e. antigenomic sequences, inactive parts of the genome, etc).

Depending what you are planning to do with the microarray data you may or may not want to utilize the normalization routines built into APT. I suggest you look into RMA (APT), SCAN (Bioconductor), and COMBAT (Bioconductor), but again this is heavily dependent on what you plan to do with the data.