Alazygatr • algatr

library(algatr)

Install all of the required packages:

alazygatr_packages()

Are you a lazy algatr? The do_everything_for_me() function runs all six landscape genomic methods in the algatr package (wingen, TESS, GDM, MMRR, RDA, and LFMM) and provides limited options for customizability. This function primarily exists for fun and to demonstrate that algatr really can be run using on a vcf and sampling coordinates; we do not encourage researchers to actually perform analyses on their data using this function!

The main arguments within this function are simple: vcf specifies the vcf, coords specifies the sampling coordinates, and envlayers specifies the environmental layers (not required). As usual, make sure your samples are in the same order between your data file and coordinates file and that your CRS is consistent!

First, let’s load our test data. For the sake of things running quickly, let’s run this function on only 20 individuals from the test dataset.

load_algatr_example()

## 
## ---------------- example dataset ----------------
##  
## Objects loaded: 
## *liz_vcf* vcfR object (1000 loci x 53 samples) 
## *liz_gendist* genetic distance matrix (Plink Distance) 
## *liz_coords* dataframe with x and y coordinates 
## *CA_env* RasterStack with example environmental layers 
## 
## -------------------------------------------------
##

##

gen <- liz_vcf[, 1:21]

## Loading required package: vcfR

## 
##    *****       ***   vcfR   ***       *****
##    This is vcfR 1.15.0 
##      browseVignettes('vcfR') # Documentation
##      citation('vcfR') # Citation
##    *****       *****      *****       *****

coords <- liz_coords[1:20, ]
envlayers <- CA_env

Now, let’s run the function:

lazy_results <- do_everything_for_me(gen, coords, envlayers, quiet = FALSE)

var	estimate	p	95% Lower	95% Upper

geodist	0.37	0.01	0.26	0.48
Intercept	0.00	0.53	−0.11	0.11

R-Squared:	0.14
F-Statistic:	29.83
F p-value:	0.01

predictor	coefficient

Geographic	0.66
CA_rPCA1	0.00
CA_rPCA2	0.31
CA_rPCA3	0.00

% Explained:	24.52¹
¹ The percentage of null deviance explained by the fitted GDM model.

r	p	snp	var
0.61	0.00	Locus_1841	CA_rPCA3
-0.58	0.00	Locus_1319	CA_rPCA2
-0.58	0.00	Locus_2394	CA_rPCA2
-0.53	0.00	Locus_1247	CA_rPCA2
-0.52	0.01	Locus_2733	CA_rPCA2
-0.51	0.01	Locus_2602	CA_rPCA2
-0.49	0.01	Locus_3143	CA_rPCA2
-0.48	0.01	Locus_2665	CA_rPCA3
0.47	0.01	Locus_350	CA_rPCA3
0.46	0.01	Locus_2400	CA_rPCA3

snp	variable	B¹	z-score	p-value	calibrated z-score	calibrated p-value	adjusted p-value
Locus_90	CA_rPCA1	0.24	4.36	0	28.07	0	0.00
Locus_1278	CA_rPCA2	-0.21	-4.89	0	15.27	0	0.01
Locus_1318	CA_rPCA2	-0.21	-4.89	0	15.27	0	0.01
Locus_1801	CA_rPCA2	-0.21	-4.89	0	15.27	0	0.01
Locus_2045	CA_rPCA2	-0.21	-4.89	0	15.27	0	0.01
Locus_2947	CA_rPCA2	-0.21	-4.89	0	15.27	0	0.01
Locus_1724	CA_rPCA2	-0.20	-4.27	0	11.62	0	0.04
Locus_1247	CA_rPCA2	-0.19	-5.57	0	19.76	0	0.01
Locus_1477	CA_rPCA2	0.17	4.70	0	14.06	0	0.01
Locus_786	CA_rPCA2	-0.17	-4.22	0	11.36	0	0.04
¹ LFMM effect size

The do_everything_for_me() function returns a list with each analysis’s results as objects named according to the method. These formats are identical to those obtained when running respective “do_everything” functions (e.g., tess_do_everything() function output is identical to that within the tess object).

For methods with model selection (e.g., GDM and MMRR), the default of do_everything_for_me() is to run with model selection ("best"), and if no significant variables are found, the function will revert to running the "full" model for these methods.