Association Analysis with Plink error

Question

[mmukhame321]

1. this is my phenotype file (called outputfile.txt in command line use):

FID IID Cadmium_Chloride    Caffeine    Calcium_Chloride    Cisplatin   Cobalt_Chloride Congo_red   Copper  Cycloheximide   Diamide E6_Berbamine    Ethanol Formamide   Galactose   Hydrogen_Peroxide   Hydroquinone    Hydroxyurea Indoleacetic_Acid   Lactate Lactose Lithium_Chloride    Magnesium_Chloride  Magnesium_Sulfate   Maltose Mannose Menadione   Neomycin    Paraquat    Raffinose   SDS Sorbitol    Trehalose   Tunicamycin x4-Hydroxybenzaldehyde  x4NQO   x5-Fluorocytosine   x5-Fluorouracil x6-Azauracil    Xylose  YNB YNB:ph3 YNB:ph8 YPD YPD:15C YPD:37C YPD:4C  Zeocin
A01_01 A01_01 -7.32351970578731 0.279992827000249 0.313118165836545 1.65817907082079 -1.60444210190495 5.84161725611811 -4.13094977046224 0.821226166664529 3.62260156257758 -0.378746805086589 -0.6449544101999 0.736772421684145 1.46869950807288 4.25247880427656 -0.439429122584143 0.471260934436784 -0.502023574403563 -0.0196386553492135 -0.520403819717771 -3.04250228422253 -0.239535833991348 3.24339670861968 -3.94506679134117 2.13462934930907 2.02778180052776 -10.930132784538 1.5331378908103 -0.768634428150619 0.718639222878471 NA -0.734761808299035 0.760529220008652 -0.756192366531865 2.09460844047308 0.20839083641332 1.39503843403223 1.19905393883646 -0.309148758204671 17.470821887375 0.055225386257017 -0.184268373327551 24.5489707854467 0.712171057826513 0.890841948461777 4.11837231021474 8.59281835912838
A01_02 A01_02 -8.09823582391425 -0.206326076018097 -0.534843782803465 -0.918011723216776 0.892197592923579 -1.61817232545715 1.13194737114694 -0.764735687307454 -2.94627867266571 -2.47519275599105 -0.203037737638922 -0.661085887535845 1.74459605348331 -3.83556423753907 -0.120208249207331 -1.99993926169807 -1.09816332010706 0.649474778852619 -0.586994384784721 2.64012099474669 -0.308361579587721 -1.14413156235224 -5.39735154446785 0.319899854554594 -1.72754411262877 5.08769611065373 -0.691267578117329 2.46423743955996 -0.706985029180143 NA -0.3904578652155 -0.598586077448249 2.3069619915968 -3.57657131625641 -0.53930290771097 0.631594704977842 0.44318172585355 0.697908024887215 18.0529250220805 0.28346211824195 1.66220146506514 26.808475766906 -1.52249804709209 0.0060616411700553 0.0665150029109814 -4.22047646027812

2. I would like to perform quantitative association testing for all traits. I have been fighting all day and cant find an answer anymore, why I am getting an error (see below), when running this command line:

$ plink2 --assoc --bfile binary_fileset --pheno outputfile.txt --pheno-name Caffeine -pfilter 1e-5

Error:

4096 MB RAM detected; reserving 2048 MB for main workspace.
11623 variants loaded from .bim file.
1008 people (0 males, 0 females, 1008 ambiguous) loaded from .fam.
Ambiguous sex IDs written to plink.nosex .
0 phenotype values present after --pheno.
Using 1 thread (no multithreaded calculations invoked).
Before main variant filters, 1008 founders and 0 nonfounders present.
Calculating allele frequencies... done.
11623 variants and 1008 people pass filters and QC.
Note: No phenotypes present.
Warning: Skipping --assoc/--model since less than two phenotypes are present.

[Philipp Bayer]

Have you tried running it with --all-pheno instead of --pheno-name Caffeine? Maybe it doesn't properly pick up the caffeine phenotype for some weird reason

Answer 1

[yulin1988320]

Hi mmukhame321, I guess you have solved this problem. It took me a while to find out why. My .fam file does not include the gender information, so there will be a warning like "If you don't want those phenotypes to be ignored, use the --allow-no-sex flag". I suppose that's why it came out to the "Note: No phenotypes present." After I added the flag "--allow-no-sex", it worked. Hope this information could help others who are disturbed by this annoying problem.

[Diesel]

Many many thanks for this. I was about to waste my full day (already 2 hours wasted) trying to understand why it was not working. It is indeed true, if the .fam does not include gender information you need to use the --allow-no-sex flag else phenotypes will be ignored.

[Kevin Blighe]

I have now toggled this old answer as accepted

Answer 2

[fanqianyin]

For my experience, the problem could be: The family ID and individual ID in your phenotype file does not match these columns in the .fam file. For example, one individual in .fam with:

head -1 yourfile.fam #view first row of the file

#FID IID phenotype1 (actually, .fam file may not contain this title row; and here I just show first three columns.)
0 individual1 2

but the same individual in .pheno (phenotype file) with:

head -2 yourfile.pheno #view first  two rows of the file

FID IID phenotype1
individual1 individual1 2

The solution is to replace "individual1" in FID column with "0" in your phenotype file, result in:

head -2 yourfile.pheno #view first  two rows of the file

FID IID phenotype1
0 individual1 2

(Note: if you use the title in the first row, i.e. "FID IID phenotype1" as above, do not forget to use flag --pheno-name phenotype1.)

Answer 3

[rms2jg]

Hi, this is an old thread by now but I just wanted to add to this post because it helped a problem I was having and hopefully it can help anyone else who finds themselves on this post. Fanqianyin's post was at least part of the answer. The other piece was I had named my samples using an underscore. It appears plink parses underscores in some special way in some commands and recommends against using them in FIDs and IIDs, which appears to be the case in your phenotype file.

Several PLINK commands (e.g. --cluster) merge the FID and IID with an underscore in their reports; for example, a sample with FID = 'Chang' and IID = 'Christopher' would be referenced as 'Chang_Christopher'. We preserve this behavior for backwards compatibility, so you should avoid using underscores in FIDs and IIDs (consider '~' instead).

-https://www.cog-genomics.org/plink/1.9/formats#fam

Renaming my samples without underscores allowed me to proceed.