rojaff / langen_pipeline

Suit of R-scripts to perform landscape genomic analyses

LanGen_pipeline

These suit of scripts were developed to perform the following analysis:

1. SNPs dataset filtering
2. Assessment of population genetic structure
3. Selection of environmental data
4. Assessment of fine-scale spatial genetic structure
5. Identification of outlier (adaptive) loci using pcadapt
6. Identification of genotype-environment associations using Redundancy Analyses (RDA) & Mahalanobis distances
7. Identification of genotype-environment associations using Latent Factor Mixed Models (LFMM2)
8. Plotting detections with a Venn Diagram
9. Visual assessment of genotype-environment associations using Redundancy Analyses (RDA) on the candidate loci
10. Mapping adaptive genetic variation using Spatial Principal Components (sPCA)
11. Predicting adaptive genotypes from RDA models and environmental data

langen_pipeline's Issues

Error in cor(x, snp.gen) : 'x' must be numeric (Loop)

Hello everyone,

I have a simple problem with this loop in the section (Correlations of each candidate SNP with the environmental predictors):

foo <- matrix(nrow=(ncand), ncol=4) # 4 columns for 4 predictors
colnames(foo) <- c("ID","POP","SURFACE","ALTITUDE")

for (i in 1:length(cand$snp)) {
nam <- cand[i,2]
snp.gen <- gen.imp[,nam]
foo[i,] <- apply(pred,2,function(x) cor(x,snp.gen))
}

#Error in cor(x, snp.gen) : 'x' must be numeric

I understand that cor is a correlation, but in the x section what kind of input should I add to the loop in this case?

Thank you very much for your help