Make Tabula Muris run steps for on aws more clear and reproducible

I have to re-run the tabula muris data, since I accidentally ran it with human transcriptome index instead of mouse transcriptome index. In my re-running it, I am finding I have encountered problems and my script is not super easy to follow for the aws part. Although for now my focus is getting the data processed, I also need to go back and get `tab_mur_data_process.sh` in better working order.