input data

[mohammadzarrabian]

Hello,

I want to use treeWAS for Bacterial GWAS, and my data is in the form of SNPs and gene presence /absence, but when I run the program I get the (Error in tree.reconstruct(genotype, method = "nj", dist.dna.model = "JC69", : There must be no more than 4 unique values in DNA, excluding NAS) I also used Nucleotide and used the code in your program (tree. reconstruct) to draw the tree and introduced it as the input of the tree, but I always get the (Phen must have names corresponding to tree$tip.label) however I check my data and couldn't find any difference in label in my phenotype and genotypic data could you please help me. Best, Mohammad

[LudoPoire]

Hi Mohammad, I know this topic is pretty old but did you find a solution to this problem ? I'm doing similar analyses. Thanks !

[caitiecollins]

Hi, If either of you could send me your input data, I'm happy to try to figure out what the issue is. Would you be able to email your snps, phen to me, with any other inputs/parameters you're using?

• CC

[caitiecollins]

My email is caitiecollins@gmail.com

[caitiecollins]

If there are privacy concerns, you could also send me a re-named subset of your data, or anything that has similar contents and reproduces the error. Best, Caitlin.

[LudoPoire]

Hi Catie, thanks for the quick reply. I was able to fix my problem by re-organizing my data and renaming my loci with numbers (1, 2, 3, ...) instead of strings (originally they were named V1, V2, etc. as they're the output of the pangenome command of PanACoTA).

But I'm now facing another problem, unrelated to this : My phenotype data is a score from 0 to 4, so I tried "phen.type =" categorical, but it yielded the following error :

• Error in chisq.test(x = Pd, y = Sd[, e], correct = F) :
• 'x' and 'y' must have at least 2 levels

Do you have a solution for this ? Should I use the "discrete" phenotype option ?

Thanks ! :)