variant calling using data from HaloPlex HS

Question

[SJ Basu]

Hello,

I have recently received HaloPlex HS data for 20 cancer samples. I was to call single nucleotide variants from the fastq reads. I have used GATK best practice pipeline to call variants and then annotate with SNPeff and annovar.

But a friend of mine, suggested that HaloPlex HS reads needs to aligned according to same index sequence then a consensus sequence which in turn should go for alignment. Though I was not clear about the idea but my question is : Does HaloPlex HS data indeed needs any special pre-alignment analysis [to take care of HS portion] to get low occurring mutations with more confidence ?

[P.S. I couldn't find any paper or company manual regarding this idea.]