SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
how to do batch correction shirley47162928 Bioinformatics 6 01-16-2015 02:05 PM
SVA input files srd Introductions 0 03-16-2011 07:17 AM
SVA: Causal variants for metachondromatosis are identified. Dongliang Ge Bioinformatics 0 06-18-2010 08:17 AM

Reply
 
Thread Tools
Old 07-27-2015, 01:25 PM   #1
shirley47162928
Member
 
Location: in

Join Date: Jan 2015
Posts: 15
Default sva package for batch correction

Hello,
I am trying to do Batch correction using SVA package (with ComBat function) for my RNA-seq raw counts data.
When performing step 4: Applying the sva fuction, I found an error message like this:

PHP Code:
n.sv num.sv(edata2,mod,method="leek")
n.sv
[114
svobj sva(edata2,mod,mod0,n.sv=n.sv)
Number of significant surrogate variables is:  14 
Iteration 
(out of 5 ):Error in density.default(xadjust adj) : 'x' contains missing values
In addition
Warning message:
In pf(fstatsdf1 = (df1 df0), df2 = (df1)) : NaNs produced
Error during wrapup
cannot open the connection 
Because my raw counts data have many genes show zero expression, I then removed those lines with zero. And I also removed batches that had only one sample each. Then this error message was still there. But I can go ahead with step 7: Applying the ComBat function and get the adjusted data.

I wonder if it correct that I remove those lines with zero.

my original expression data looks like this:
PHP Code:
head(edata)
         
FH1.B2.K100813 FH0.B3.K100823 FH1.B2.K100826 FH1.B2.K100831 FH0.B2.K101448 FH0.B2.K102654 FH0.B5.K104200 FH0.B5.K104238 FH0.B2.K104250 FH0.B2.K104343 FH0.B3.K104403 FH0.B5.K104443
40969             69.00          11.00           4.00          55.00          78.67         138.00         109.00          70.00         144.00         127.92         118.00          57.00
40970            569.43          73.22           6.00         534.51         153.15         400.14         351.60         440.97         127.28         214.97         106.65         455.01
40974           1647.12         150.30          51.58        1932.85        1018.62        1039.99        2019.72        2080.37        1634.20         847.77         882.49        1834.79
ADAM23             5.01           0.00           0.00           4.96          32.00           0.00          33.00          19.91          30.01          10.01          23.00           9.92
ASNS              91.39          31.00           6.00          99.32         157.44         130.71         176.98          76.58         290.99         113.38         127.00         101.89
C14ORF10          14.01           4.07           0.00          32.78          15.74           8.29          22.71          31.55          19.09          18.76           7.58          19.44
         FH1
.B5.K104506 FH0.B2.K104603 FH0.B2.K104638 FH1.B2.K104824
40969            120.00         191.00          56.00          58.00
40970            550.40         764.77         310.11         329.31
40974           1742.01        2296.24        1081.09         911.42
ADAM23            19.91          39.98          20.02           4.98
ASNS              84.12          89.83          63.28          95.25
C14ORF10          13.40          29.57          15.75          11.14 
shirley47162928 is offline   Reply With Quote
Old 08-24-2016, 02:32 PM   #2
boilermaker
Junior Member
 
Location: Ohio

Join Date: Apr 2014
Posts: 5
Default Did you find the solution?

I am getting the same error message. Did you ever find an answer?
boilermaker is offline   Reply With Quote
Old 09-28-2016, 09:48 AM   #3
fp1
Junior Member
 
Location: Cambridge

Join Date: Nov 2014
Posts: 1
Default Also having the same problem...

...and wondering if you found an answer?

In my case, the number of significant surrogate variables estimated = 6

I do find that everything runs fine if I lower that to 5, but I'm really not sure why that should be...
fp1 is offline   Reply With Quote
Old 10-18-2017, 10:48 PM   #4
anton.kratz
Junior Member
 
Location: San Diego, CA, USA

Join Date: Jun 2016
Posts: 7
Default

I am running into the very same error:

"Error in density.default(x, adjust = adj) : 'x' contains missing values"

I found various posts (stackoverflow, bioconductor support forum, biostars) referrin to this error but now solution, only workarounds without any understanding or explanation what causes the error, or why workarounds seem to work. Suggested solutions were: reducing the number of n.sv until it works, removing genes with low counts until it works.

Did anyone find a solution to this?
anton.kratz is offline   Reply With Quote
Reply

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 10:00 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO