More linear pipeline structure for supporting multiple file types

[adamrtalbot]

This PR changes the pipeline structure so that it uses a more linear set of steps and can be controlled.

Changes:

• Create FASTQ and create files composed of zeros are performed at the start of the pipeline
• They are fed into subsequent processes for file manipulations
• The two file types are manipulated together but can be switched on or off based on flags

Advantages:

• This is a more flexible structure which means we can extend it to new file types when they are necessary
• We can perform the file manipulations on larger or smaller files by enabling or disabling features
• We have granular control over more aspects of the pipeline

Disadvantage

• Separate file operations cannot be performed on each file type, since they are mixed. We can disable individual processes though.
• It may be necessary to launch multiple pipelines to test multiple things instead of one test with multiple hypotheses
• If statements (yuck)

Supercedes #3, closes #4