lcMLkin (Maximum Likelihood Estimation of Relatedness)

How to Compile:

Note: This code requires a C++11 compliant compiler. g++ >= 4.7.3 and clang >= 3.4 should be OK

1.) > git clone https://github.com/COMBINE-lab/maximum-likelihood-relatedness-estimation.git

2.) > cd maximum-likelihood-relatedness-estimation

3.) > make

Now, you should have an executable in this directory called lcmlkin. You can run it with the -h option to get help. A typical run would look something like:

4.) > lcmlkin -i input.vcf -o output.relate -g all -t 8

This will run lcmlkin on the file input.vcf, estimate background allele frequencies assuming all individuals are unrelated, and using the variant of the algorithm that sums over all genotypes given their likelihoods (-g all). It will write the output to the file output.relate, and will make use of up to 8 worker threads (-t 8).

Producing VCF input from BAM files

A preliminary implementation of a script to generate lcMLkin-ready VCF files from a set of indexed BAM files is available under the src_python/SNPbam2vcf directory of the repository (here). There, you will find more information on the script including usage examples.

Best Practices

Detailed information on how to prepare input files with likelihoods and how to choose SNPs for analysis can be found in our [wiki] (https://github.com/COMBINE-lab/maximum-likelihood-relatedness-estimation/wiki).