| | AdapterRemoval | Trimmomatic |
|---|
| | Paired-end | Single-end | Paired-end | Single-end |
|---|
Run time (s) | 139.42 | 36.07 | 39.9 | 20.1 |
Memory (kB) | 7,664 | 6,512 | 387,488 | 611,536 |
Trimmed, no adapter | 95 | (0.02%) | 93,909 | (15.00%) | 0 | (0.00%) | 0 | (0.00%) |
Trimmed exact | 346,806 | (92.74%) | 327,925 | (87.69%) | 120284 | (32.16%) | 120,284 | (32.16%) |
Trimmed more | 0 | (0.00%) | 298 | (0.08%) | 0 | (0.00%) | 0 | (0.00%) |
Trimmed less | 0 | (0.00%) | 1,038 | (0.28%) | 0 | (0.00%) | 0 | (0.00%) |
Missed adapter | 27,157 | (7.26%) | 44,702 | (11.95%) | 253679 | (67.84%) | 253,679 | (67.84%) |
PPV | 1.00 | 0.78 | 1.00 | 1.00 |
SEN | 0.93 | 0.88 | 0.32 | 0.32 |
SPEC | 1.00 | 0.85 | 1.00 | 1.00 |
MCC | 0.94 | 0.71 | 0.48 | 0.48 |
- Statistics for AdapterRemoval and Trimmomatic tested on both single-end and paired-end data. The test set contains 1,000,000 read pairs of which 373,963 contain adapter fragments and the remaining 626,037 do not. The paired-end test was performed on the 1,000,000 pairs, and the single-end test was performed on the first read from each pair. The reported numbers are: Run time (seconds); max memory usage (kilobytes); number of reads containing no adapter that were trimmed; number of reads with adapter where just the adapter was removed; number of reads with adapter where more than the adapter was trimmed; number of trimmed reads with adapter where the full adapter was not removed; number of reads with adapter where no trimming was performed; positive predictive value; sensitivity; specificity; Matthew’s correlation coefficient. For “Trimmed, no adapter”, the percentage of the 626,037 reads with no adapter that were trimmed is shown. For the following four rows, the percentages are of the 373,963 read pairs with adapter, and these numbers add up to 373,963 and 100%, respectively.
