I am working on phasing vcf files using EagleImp-Web, a web tool that uses Eagle2 and a reference panel (1000 genomes or HRC) to phase variants per chromosome. For my research, I only need about 100 variants split among 10 chromosomes to be phased. Does phasing effectiveness rely on the number of variants included within a vcf file? I could provide a vcf with all variants within each chromosome, then filter out the ones I need. Or a vcf with just the variants I need for my research, which is much faster. I am confused about how reference panel Eagle2 phasing works.