encode gt of hg38 to machine learning

Question

[أرْوَى]

Hi

I'm new in the field

I have a large vcf file that have many variants with many samples. I extract gt for each sample / variants to get a matrix to do a machine learning algorithm. Now I need to encode this gt to do a machine learning.

I see a stranger numbers in gt like [-1 -1] , [1 5] , [ 1 6 ] , [ 0 -1] ,[ 1 -1],[ 2 -1]

So, what is mean for hg38? and How can I encode to use a machine learning?

[Pierre Lindenbaum]

I see a stranger numbers in gt like [-1 -1] , [1 5] , [ 1 6 ] , [ 0 -1] ,[ 1 -1],[ 2 -1]

where do you find this information in the VCF ? how do you extract the matrix ?

[أرْوَى]

I'm extracting gt using sckite allel package ( python ). when I compare what I actually extract and origenal vcf I found ( -1 = . ).