refactor: improve/extend functionality of runDoseResponseNgsPipeline

Details:

user can select the step from which the pipeline can be run
user can select experiments for which the pipeline will be run
user can run the pipeline with key input provided as a data.frame (previously) or MAE (new)
added logic for saving/reading (per experiment/step) intermedia data
added helper functions:
- get_pipeline_steps
- do_skip_test
- is_preceding_step
- save_intermediate_data
- read_intermediate_data
- get_data_type_to_data_model_mapping
- validate_data_models_availability
'data_model' S3 methods added (MAE, data.frame)
'prepare_input' S3 methods added (MAE, data.frame)
logic for nested_identifiers updated (nested_identifiers_l vs nested_identifiers charvec)
data_type parameter introduced
create_and_normalize split into two steps (in the runDoseResponseNgsPipeline)

Checklist for the sustainable code base

[ ] I added tests for any code changed/added
[X] I added documentation for any code changed/added
[X] I made sure the naming of any new functions is self-explanatory and consistent

Logistic checklist