Improve compatibility between pipeline & cluster

[Cecilia-Sensalari]

• Removed $NSLOTS in main.nf, which is defined only by SGE; in its place, $task.cpus is used
• Added Nextflow directives in configuration files such as cpus, penv, memory
• Updated Nextflow configuration file description in documentation

[Cecilia-Sensalari]

UPDATES Nextflow configuration files:

• Add cpus directive under executor for when executor is set to local: it limits the number of CPUs available for the executor (e.g. Docker contains would otherwise use as much resources as they can).
• Add queueSize directive under executor for when the pipeline is run on a cluster, to limit the number of processes simultaneously handled by the executor (handy when the user can't require too many CPUs at the same time).
• Removed sequential, nThreadsParalogs, nThreadsOrthologs and unnecessary maxForks parameters in main.nf
• Docs updated with description of queueSize and cpus.

[Cecilia-Sensalari]

Updates:

• Docs considers now Nextflow configuration file as a required input and not anymore as optional
• Add sentence about setting queueSize to 1 to have sequential workflow

[Cecilia-Sensalari]

UPDATES:

• Accepted review edits
• NF configuration files for test and example dataset have been renamed nextflow.config and the docs explain how to use -c option
• Readme has been updated: Nextflow config file is now mandatory, the -c option is removed, the distinction between local and remote run has been removed and I added a "Prepare the configuration files" step (such as in usage.rst).

[Cecilia-Sensalari]

Removed the last -c ./custom_nextflow.config and -with-singularity vibpsb/ksrates:latest. Removed in installation.rst the -with-container commands and mentioned nextflow.config and how to use it.