I gave a talk on Bash yesterday; you can see the slides if you're interested. Figured out that it'll take 3-4 weeks to print my lab book, which effectively means it needs to be ready to print by the last week of November, so that's where my Thanksgiving break is going to go...
You should most definitely check out https://software-carpentry.org/lessons/ -- it's targeted at non-CS people doing data science and is quite well written. Also, to toot my own horn, I have written a course on this sort of stuff: http://web.mst.edu/~nmjxv3/cs1001/
I had troubles figuring out something else. I have two input files and 4 output files per sequence and 115 of those. So I needed a for loop that takes the name of the sequence (blabla001_1.seq) finds the second part of the pair (blabla002_2.seq) then feeds that into a java command that gives 4 outputs using the names of the files. I am used to writing scripts in R so I didn't know how to do that for bash. At the end I found a script that is supposed to do it, took it apart and understood how it works. 10 minutes ago I got my results :)