- Decrease number of processed sequences
- Choose random sequence numbers to process
- Refactor test code for readability
- Use threadpool for I/O parallel processing
- Add logs instead of prints
- Only apply find-difference code on a random file.