[WIP] Feature/update for new scanpy scripts

[pinin4fjords]

This PR updates the clustering workflow taking into account changes made by Ni in updating scanpy-scripts. It retains as far as possible the previous parameterisations, which sometimes means overriding Ni's defaults. For example

• The default is now to filter marker genes etc by fold change, which I don't believe we did previously, and which I have therefore disabled here.
• We needed an additional normalisation without log-transform to generate our normalised matrix, since the new normalisation step should have log transformation to pass to the next step.

I had to restore output renamings lost on tool update/ replacement.

I have tested that this works on a small dataset on the Galaxy GUI, and it works without error via Bioblend.

Please don't merge without discussion- applying this will likely involve re-running all of our tertiary analysis.

[pcm32]

Closing to avoid confusion, replaced by #16