NINJA-OPS: Fast Accurate Marker Gene Alignment Using Concatenated Ribosomes
Fig 3
Comparison of NINJA and USEARCH 8 best match accuracy using all unique matches.
These histograms show the distribution of matches by algorithm, where NINJA match proportion is in red, USEARCH in blue, and the overlap in purple. The right-most bars in each histogram show more red than purple, indicating NINJA aligned more reads at higher accuracy than USEARCH in the corresponding bins. The thick vertical bars correspond to the average alignment accuracy, and the thin curves represent the interpolated distribution. A Student’s t-test on the mean alignment accuracy shows NINJA’s mean ID was 98.9%, while USEARCH 8 produced 98.7% with p-value < 2.2e-16. For USEARCH 8, fast settings are used (corresponding to the defaults in the QIIME pipeline). Total number of unique sequences mapped are 168,761, 170,063, and 169,313, for NINJA [default], NINJA [max], and USEARCH 8, respectively.