[GATK4.0]Which steps should be done before 'ReadsPipelineSpark '?

In the latest version GATK4.0, the function of pipeline 'ReadsPipelineSpark ' is that
Takes unaligned or aligned reads and runs BWA (if specified), MarkDuplicates, BQSR, and HaplotypeCaller to generate a VCF file of variants.
In my opinion, it means contains bwa,MarkDuplicates, BQSR, and HaplotypeCaller in one command. So which steps should be done before 'ReadsPipelineSpark '?
I try to use FastqToSam->AddOrReplaceReadGroups->ReadsPipelineSpark, but it is not correct. So what is the correct pipeline?