Table 2 |
|||
| One possible set of DNA sequence data compression factors for the various experimental classes | |||
| Class | Description | Rate for Physically Unique samples | Rate for Physically Archived/archivable Samples |
| 1 | Historical sampling of environment or time specific elements | 1.0 | 1.0 |
| 2 | Very rare objects | 1.0 | 1.0 |
| 3 | Longitudinal studies which could in theory be rerun in the future but have a > 10 year horizon to recreate | 1.0 | 2.0 |
| 4 | Samples acquired from patients or animals with a high individual acquisition cost, but a conceptually continuous generation | 1.0 | 10.0 |
| 5 | A complex experiment with > 6 month resource development | 10.0 | 100.0 |
| 6 | A routine experiment with < 6 month resource development | 20.0 | 200.0 |
| 7 | Verification experiment as a component in an overall flow | 1000.0 | ∞ (Infinite compression of data indicates no data archiving; it may, however, be useful simply to record that the experiment was carried out.) |
Compression is higher for data that are easy or inexpensive to reproduce, and lower for data derived from unique or irreproducible samples.
Cochrane et al. GigaScience 2012 1:2 doi:10.1186/2047-217X-1-2
