In this script we take two raw fluidigm (juno) datasets for Iberian lynx in Spain and Portugal and curate them for downstream analysis

Identifying candidate outlier SNPs from PLINK FST data on 6SV.vcf.gz We’ll be analyzing the fst_circ_therm_genelist_6SV.fst which has been downloaded locally our project space. This dataset includes 1650 SNPs within a 54-gene list related to circadian rhythm and thermo-tolerance in lynx. This gene set was pre-identified using primary literature

Identifying candidate outliers from population-level whole genome resequencing for Canada lynx

We apply the TropFishR package in RStudio to perform a single-species fish stock assessment with length frequency (LFQ) data. This includes the following steps: (1) estimation of biological stock characteristics (growth and natural mortality), (2) exploration of fisheries aspects (exploitation rate and selectivity), (3) assessment of stock size and status.

This is a shorter version of the prior Rpubs document that glazes over most of the code and methods. We display 23 individual estimates of admixture (K= 1:6) by overlaying pie-charts on a geographic map of the White Mountain National Forest.

We display the admixture coefficients (results from sNMF) for 23 red squirrels on a geographic map of the White Mountains National Forest using ggplot2.

This analysis was revised using a VCF file with 23 individuals instead of 24 in order to remove a duplicated individual. We reduced our 78k SNP dataset using LD-based SNP pruning. Then we confirmed the results of our sNMF analysis using a PCA, followed by two identity by descent relatedness analyses (MoM and MLE).

We performed a revised sNMP analysis using 23 individuals rather than 24. We dropped UMA4 from the vcf file because it was a duplicate of UMA21. This was accomplished using vcftools. The results indicated that the optimal number of clusters was K=2 rather than K=3 in our previous analysis.

We reduce our 78k SNP dataset using LD-based SNP pruning. Then we will confirm the results of our sNMF analysis using a PCA,followed by two identity by descent relatedness analyses (MoM and MLE)

We generated individual ancestry coefficients using sNMF, a computer program based on sparse nonnegative matrix factorization (NMF) which is a likelihood algorithm comparable to that used in STRUCTURE or ADMIXTURE. This method was selected because it is optimized for large SNP datasets and greatly reduced computation time. We used a pilot SNP dataset generated from ddRAD sequencing of 24 red squirrels distributed across Mt. Jefferson and Mt. Washington in the White Mountains National Forest.