Sambamba markdup usage help

Question

[jonas.andersson]

Hi!

I need so help regarding the usage of sambamba markdup. I have read the documentation but I don't quite understand.

1. What is meant by insert size here?

    --hash-table-size=HASH_TABLE_SIZE
                       size of hash table for finding read pairs (default is 262144 reads);
                       will be rounded down to the nearest power of two;
                       should be > (average coverage) * (insert size) for good performance
   

2. To get 100 GB here, should I just write: --sort-buffer-size 102400 ? The reason I wonder is that in sambamba sort you should specify e.g. Mb or Gb after the integer.

     --sort-buffer-size=SORT_BUFFER_SIZE
                           total amount of memory (in *megabytes*) used for sorting purposes;
                           the default is 2048, increasing it will reduce the number of created
                           temporary files and the time spent in the main thread
   

thx / Jonas