Compression by data type.

Showing the compressed file size break down by bits per sequence identifier, per base-call and per quality value. In some cases these sizes refer to cases where a reference was previously used to map, but it has not been used during compression (e.g. BAM). The ID, Base and Qual columns are the number of bits required to store the complete sequence identifier, a single base nucleotide and a single quality value respectively. The C.R. column is the compression rate in MB per second. Mem is the amount of memory required during compression. References used were human hg19 and C.Elegans WS233. Non-reference based Quip used the “-a” assembly option for high compression mode.aGoby does not store unmapped data. The Goby figures have been estimated by adding 2 bits per absent base-call and scaling up the name and quality figures by the percentage of unmapped reads.