Missing values in pheno file?

smiler76
New Member

Posts: 4

Missing values in pheno file? Oct 4, 2017 13:54:10 GMT

Quote

Post by smiler76 on Oct 4, 2017 13:54:10 GMT

I'm new to using GCTA, so any help would be appreciated. I've been trying to run mlmm-loco, but have been having trouble loading the phenotype file. I get the error message "Error: 2 phenotype values are missing in line #2 in the file [../../pathway/FileName]". I've checked through the file for missing values and cannot find problems. Can anyone give some advice?

Complete message below:

Reading PLINK FAM file from [allChr_282_onHmp321.FL06.merge.fam].
192 individuals to be included from [allChr_282_onHmp321.FL06.merge.fam].
Reading PLINK BIM file from [allChr_282_onHmp321.FL06.merge.bim].
976308 SNPs to be included from [allChr_282_onHmp321.FL06.merge.bim].
Reading PLINK BED file from [allChr_282_onHmp321.FL06.merge.bed] in SNP-major format ...
Genotype data for 192 individuals and 976308 SNPs to be included from [allChr_282_onHmp321.FL06.merge.bed].
Reading phenotypes from [../../../GWASdatasets/FL06/1.meanByLineandLoc.divpanel.LocSpecificResids.txt].
There are 21 traits specified in the file [../../../GWASdatasets/FL06/1.meanByLineandLoc.divpanel.LocSpecificResids.txt].
Trait #1 is included for analysis.

Error: 2 phenotype values are missing in line #2 in the file [../../../GWASdatasets/FL06/1.meanByLineandLoc.divpanel.LocSpecificResids.txt]

Thanks

xue New Member Posts: 3	Missing values in pheno file? Oct 4, 2017 20:21:54 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by xue on Oct 4, 2017 20:21:54 GMT There is something wrong in the 2-line of your phenotype file. Maybe you use "." (or other characters) as your missing value.In gcta,missing value should be represented by "-9" or "NA".
	Last Edit: Oct 4, 2017 20:31:27 GMT by xue

smiler76
New Member

Posts: 4

Missing values in pheno file? Oct 4, 2017 21:16:25 GMT

Quote

Post by smiler76 on Oct 4, 2017 21:16:25 GMT

Thanks for the reply. The txt file was created in R, so all missing values are scripted to be "NA". I have normalized the data by using linear model residuals, so some of the values are negative. Could negative values cause a problem?

xue New Member Posts: 3	Missing values in pheno file? Oct 5, 2017 1:25:13 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by xue on Oct 5, 2017 1:25:13 GMT I don't know the reason now, I think you can give a screenshot of your phenotype data, or you can just use head command to get first few lines to show the specific problems.

smiler76
New Member

Posts: 4

Missing values in pheno file? Oct 5, 2017 15:38:25 GMT

Quote

Post by smiler76 on Oct 5, 2017 15:38:25 GMT

So, here is a reduced phenotype file. It only contains the first phenotype. The original file was 2100 phenotypes. However, I still get the same error on this as well.

> head(phenos)
Pedigree loc weight
1 282set_33-16 FL06 299.8285
2 282set_4722 FL06 155.5936
7 282set_A6 FL06 296.5623
3 282set_A188 FL06 252.6267
4 282set_A441-5 FL06 329.8363
5 282set_A554 FL06 106.7866

my PED (.fam) file looks like this:

> head(fam.file)
V1 V2 V3 V4 V5 V6
1 282set_33-16 282set_33-16 0 0 0 -9
2 282set_4722 282set_4722 0 0 0 -9
3 282set_A6 282set_A6 0 0 0 -9
4 282set_A188 282set_A188 0 0 0 -9
5 282set_A441-5 282set_A441-5 0 0 0 -9
6 282set_A554 282set_A554 0 0 0 -9

Last Edit: Oct 5, 2017 16:41:22 GMT by smiler76

pvdmost
New Member

Posts: 5

Missing values in pheno file? Oct 6, 2017 13:08:51 GMT

Quote

Post by pvdmost on Oct 6, 2017 13:08:51 GMT

Oct 5, 2017 15:38:25 GMT smiler76 said:

So, here is a reduced phenotype file. It only contains the first phenotype. The original file was 2100 phenotypes. However, I still get the same error on this as well.

> head(phenos)
Pedigree loc weight
1 282set_33-16 FL06 299.8285
2 282set_4722 FL06 155.5936
7 282set_A6 FL06 296.5623
3 282set_A188 FL06 252.6267
4 282set_A441-5 FL06 329.8363
5 282set_A554 FL06 106.7866

my PED (.fam) file looks like this:

> head(fam.file)
V1 V2 V3 V4 V5 V6
1 282set_33-16 282set_33-16 0 0 0 -9
2 282set_4722 282set_4722 0 0 0 -9
3 282set_A6 282set_A6 0 0 0 -9
4 282set_A188 282set_A188 0 0 0 -9
5 282set_A441-5 282set_A441-5 0 0 0 -9
6 282set_A554 282set_A554 0 0 0 -9

Not sure if this is the cause, but in that fenotype file I see three column names, but four columns. Are the "FL06" values part of the individual ID? The FAM file does without them.

smiler76
New Member

Posts: 4

Missing values in pheno file? Oct 12, 2017 20:16:47 GMT

Quote

Post by smiler76 on Oct 12, 2017 20:16:47 GMT

Thanks for your responses. I eventually worked it out. There were several things that were wrong with the file. After using cbind() in R to create the file it added invisible quotes to the first two columns. I removed those. The second thing was that it didn't work when I used comma-, or tab- separated values in the .txt file. It only worked when I separated the values with a space. Following that, I removed the column headers, and then, everything worked.

Post by smiler76 on Oct 4, 2017 13:54:10 GMT

Post by xue on Oct 4, 2017 20:21:54 GMT

Post by smiler76 on Oct 4, 2017 21:16:25 GMT

Post by xue on Oct 5, 2017 1:25:13 GMT

Post by smiler76 on Oct 5, 2017 15:38:25 GMT

Post by pvdmost on Oct 6, 2017 13:08:51 GMT

Post by smiler76 on Oct 12, 2017 20:16:47 GMT