Should I assume that the genotype of the 1000 Genomes project individuals in positions that are not in the vcf file, are homozygous for the reference allele?

Question

[magda]

I want to merge my variant data with those from the 1000 genomes project, but which genotype can I assume that the 1000 Genomes project individuals have in the positions that are not included in their vcf files. Should I assume that they are homozygous for the reference allele? But what if in that position an individual don't have any reads? Do you know if they provide a g.vcf file, with information in all positions (reads, etc..., which I haven't found. Is there a way of extracting the genotype without having to apply the whole pipeline in the Bam files they provide? What do you do?

Thank you Magda