Review

[bgpalmer]

Looks good! I rewatched the lecture and your duplicate checks seem to add up. Specifically,

- Create tuple containing read_umi, bit16_flag, chromosome_name, adjusted_starting_position, and assign it to "key" variable. - Note: If bitwise flag of one read and bitwise flag of another read have the same bit-16 flag states (either BOTH have bit16_flag == TRUE, or BOTH have bit16_flag == FALSE), then they both have the same strandedness and possibly could be duplicates.

I also like how you export the data into the output sam file once moving on to the next umi. Sorting the file (UMI, chromosome, read position) before hand was a good call so you could implement this.