SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
p-value and log fold change Bidfudge Bioinformatics 3 07-25-2016 04:00 AM
RPKM vs Fold Change ppm Bioinformatics 4 05-06-2014 08:49 AM
Log fold change Erwan Bioinformatics 3 03-25-2014 06:05 AM
Standardized Fold Change dbaud Introductions 5 01-27-2014 11:47 AM
fold change value-cuffdiff madsaan Bioinformatics 4 02-10-2011 07:51 AM

Reply
 
Thread Tools
Old 08-16-2016, 10:56 AM   #1
granger
Junior Member
 
Location: United Kingdom

Join Date: Mar 2011
Posts: 6
Unhappy Help interpreting this fold change plot

I came across this FC plot showing "agreement" between the Nanostring nCounter Single Cell Gene Expression method and the normal (multicell) one.
http://www.nanostring.com/products/single_cell



"These data show a 1:1 correlation of fold changes between assays, demonstrating the simultaneous, unbiased amplification of hundreds of target transcripts and preservation of fold change information provided by nCounter Single Cell Gene Expression protocol."

I have a few issues with this conclusion and I'm wondering if you agree with them or am I missing something:
1) The axes have 0, which means it's probably logFC as opposed to the labels?
2) The upper left section has genes with opposite fold changes in the two techniques. This is about 1/3 of the total!
3) Even if we have a cutoff and consider only fold changes above |2| (log or otherwise), then still the two techniques do not agree on the majority of the genes. (see 0 -- -2 (or -1) band for single cell, and 0 -- 2 band for GX assay)
granger is offline   Reply With Quote
Old 08-16-2016, 12:07 PM   #2
dpryan
Devon Ryan
 
Location: Freiburg, Germany

Join Date: Jul 2011
Posts: 3,479
Default

1) Yup, they must mean logFC.
2) Yeah, but this is due to an apparent constant offset, so if real it's at least easy to correct for.
3) Take the offset into account and they're in nice agreement.

What they need to answer is why their single-cell data seems biased by a log2 of +1.
dpryan is offline   Reply With Quote
Old 08-22-2016, 12:30 PM   #3
granger
Junior Member
 
Location: United Kingdom

Join Date: Mar 2011
Posts: 6
Default

Thanks. So either the single-cell protocol failed to properly amplify the brain reference (denominator) and/or overamplified the human ref (nominator) compared to the original protocol.
Alternatively, the original protocol had these problems. And that's my problem with the correction, you can't tell which should be corrected.

And even so, we are still left with huge variation.

I think they should simply correlate the count numbers (and show on two plots).

Last edited by granger; 08-22-2016 at 12:41 PM.
granger is offline   Reply With Quote
Reply

Tags
nanostring, ncounter, single cell

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 01:28 PM.


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